<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
        <title>TheRouter.ai — Latest AI News</title>
        <link>https://therouter.ai/news/</link>
        <description>Curated AI news, model releases, and research summaries from TheRouter.ai.</description>
        <lastBuildDate>Sun, 11 Oct 2026 01:02:21 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>en-US</language>
        <copyright>© 2026 TheRouter.ai</copyright>
        <atom:link href="https://therouter.ai/news/rss.xml/" rel="self" type="application/rss+xml"/>
        <item>
            <title><![CDATA[Claude Haiku 5.5 Ships Effort Controls and Tiered Context Pricing for High-Volume Subagent Work]]></title>
            <link>https://therouter.ai/news/claude-haiku-5-5-effort-controls-subagent/</link>
            <guid>https://therouter.ai/news/claude-haiku-5-5-effort-controls-subagent/</guid>
            <pubDate>Sun, 11 Oct 2026 01:02:21 GMT</pubDate>
            <description><![CDATA[Anthropic's Claude Haiku 5.5 is the first Haiku with effort controls, priced at $0.10 per million input tokens for prompts up to 100K tokens—built for classification, subagent delegation, and real-time workloads at scale.]]></description>
            <category>industry</category>
            <enclosure url="https://therouter.ai/news/claude-haiku-5-5-effort-controls-subagent/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Sonnet 5.5 Ships Five Breaking API Changes for Operators Running Sonnet 5]]></title>
            <link>https://therouter.ai/news/claude-sonnet-5-5-five-breaking-changes-api-operator/</link>
            <guid>https://therouter.ai/news/claude-sonnet-5-5-five-breaking-changes-api-operator/</guid>
            <pubDate>Mon, 05 Oct 2026 11:22:36 GMT</pubDate>
            <description><![CDATA[Claude Sonnet 5.5 (Sep 28) ships 5 breaking changes: forced tool use returns 400, `thinking:disabled` is rejected, thinking blocks are model-bound, `computer_20251124` is gone on the Claude API, and advisor pairings are now 5.x-only.]]></description>
            <category>anthropic</category>
            <category>migration</category>
            <category>api</category>
            <category>routing</category>
            <enclosure url="https://therouter.ai/news/claude-sonnet-5-5-five-breaking-changes-api-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[A Model-Level 429 From Zhipu Took GLM Flash Models Offline. TheRouter Now Scopes Cooldowns Per Model and Gives Them a Second Route.]]></title>
            <link>https://therouter.ai/news/zhipu-flash-zai-fallback-legs-429-routing/</link>
            <guid>https://therouter.ai/news/zhipu-flash-zai-fallback-legs-429-routing/</guid>
            <pubDate>Sat, 03 Oct 2026 14:12:57 GMT</pubDate>
            <description><![CDATA[On 2026-10-02 Zhipu's API returned a model-level rate-limit response for glm-4.6v-flash and TheRouter parked the whole account. Cooldowns are now per model, and six GLM flash-family models have a second route through Zhipu's international service.]]></description>
            <category>zhipu</category>
            <category>routing</category>
            <category>incident</category>
            <category>fallback</category>
            <enclosure url="https://therouter.ai/news/zhipu-flash-zai-fallback-legs-429-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.281: Bedrock Upstreams Get Cross-Account IAM and Guardrail Enforcement]]></title>
            <link>https://therouter.ai/news/claude-code-2-1-281-bedrock-assume-role-guardrail-gateway-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2-1-281-bedrock-assume-role-guardrail-gateway-operator/</guid>
            <pubDate>Thu, 24 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[2.1.281 adds assume_role and guardrail to Bedrock upstreams. assume_role swaps long-lived IAM credentials for per-developer STS tokens. guardrail applies a Bedrock guardrail to every request. Both shift the trust boundary in multi-account AWS deployments.]]></description>
            <category>claude-code</category>
            <category>operator</category>
            <category>gateway</category>
            <category>aws-bedrock</category>
            <category>iam</category>
            <category>llm-api-ops</category>
            <enclosure url="https://therouter.ai/news/claude-code-2-1-281-bedrock-assume-role-guardrail-gateway-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Opus 5.5: Four Breaking API Changes and What They Mean for Your Routing Setup]]></title>
            <link>https://therouter.ai/news/claude-opus-5-5-breaking-changes-migration-routing/</link>
            <guid>https://therouter.ai/news/claude-opus-5-5-breaking-changes-migration-routing/</guid>
            <pubDate>Wed, 23 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Four breaking changes in Claude Opus 5.5: thinking can't be disabled, forced tool_choice returns 400, thinking blocks don't cross non-Fable/Mythos models, and computer_20251124 is gone. Each has a specific fix — three carry fallback routing implications the announcement skips.]]></description>
            <category>anthropic</category>
            <category>claude-api</category>
            <category>operator</category>
            <category>migration</category>
            <category>llm-api-ops</category>
            <category>routing-policy</category>
            <category>model-release</category>
            <enclosure url="https://therouter.ai/news/claude-opus-5-5-breaking-changes-migration-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.275 Broke Every Gateway Proxy. 2.1.276 Fixed It the Same Day.]]></title>
            <link>https://therouter.ai/news/claude-code-2275-advisor-tag-gateway-400-regression-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2275-advisor-tag-gateway-400-regression-operator/</guid>
            <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[A new internal request tag in 2.1.275 caused 400 errors on every proxy-routed API call. 2.1.276 hotfixed it the same day. Breakdown of the failure, affected configs, and three secondary operator changes worth auditing.]]></description>
            <category>claude-code</category>
            <category>operator</category>
            <category>gateway</category>
            <category>regression</category>
            <category>coding-agents</category>
            <enclosure url="https://therouter.ai/news/claude-code-2275-advisor-tag-gateway-400-regression-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI API Key Creation Governance: How Service-Account-Only Enforcement Closes the Shadow-Routing Gap]]></title>
            <link>https://therouter.ai/news/openai-api-key-creation-governance-service-account-operator/</link>
            <guid>https://therouter.ai/news/openai-api-key-creation-governance-service-account-operator/</guid>
            <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI's new org-level key creation controls let admins enforce service-account-only, project-keys-only, or disable all key creation. The first structural control that prevents shadow routing via personal API keys.]]></description>
            <category>openai</category>
            <category>governance</category>
            <category>billing</category>
            <category>operator</category>
            <enclosure url="https://therouter.ai/news/openai-api-key-creation-governance-service-account-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude API: Compaction on Demand and the `auto` Permission Mode Change Your Agentic Loop Design]]></title>
            <link>https://therouter.ai/news/claude-api-compaction-auto-permission-operator/</link>
            <guid>https://therouter.ai/news/claude-api-compaction-auto-permission-operator/</guid>
            <pubDate>Thu, 17 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Two new Claude API betas: `compact-2026-09-04` moves summarization off the critical path into a background call, and the `auto` permission mode shifts tool trust evaluation from your code to the server. Both change how you design stateful agentic loops.]]></description>
            <category>anthropic</category>
            <category>claude-api</category>
            <category>operator</category>
            <category>context-management</category>
            <category>managed-agents</category>
            <category>agentic-workflows</category>
            <category>llm-api-ops</category>
            <enclosure url="https://therouter.ai/news/claude-api-compaction-auto-permission-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.274: MCP Reliability Overhaul, Gateway Postgres Config, and Self-Healing Transcripts]]></title>
            <link>https://therouter.ai/news/claude-code-2274-gateway-postgres-timeout-transcript-recovery-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2274-gateway-postgres-timeout-transcript-recovery-operator/</guid>
            <pubDate>Thu, 17 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.274 fixes six MCP failure modes that silently break production tool sessions, adds store.connect_timeout_seconds and CLAUDE_CODE_GATEWAY_DRAIN_TIMEOUT_MS to the Claude apps gateway, and makes corrupted transcripts self-heal instead of looping forever.]]></description>
            <category>claude-code</category>
            <category>operator</category>
            <category>gateway</category>
            <category>mcp</category>
            <category>coding-agents</category>
            <category>observability</category>
            <category>reliability</category>
            <enclosure url="https://therouter.ai/news/claude-code-2274-gateway-postgres-timeout-transcript-recovery-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.273: Five New Gateway Headers and a Classifier Flip on Bedrock, Vertex, and Foundry]]></title>
            <link>https://therouter.ai/news/claude-code-2273-gateway-hint-headers-classifier-flip-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2273-gateway-hint-headers-classifier-flip-operator/</guid>
            <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.273 ships opt-in gateway hint headers exposing request class, agent type, and compaction state to any LLM proxy. It also flips the auto mode classifier to local-only on Bedrock, Vertex AI, and Foundry — only one change has a revert path.]]></description>
            <category>claude-code</category>
            <category>operator</category>
            <category>gateway</category>
            <category>coding-agents</category>
            <category>bedrock</category>
            <category>vertex</category>
            <category>observability</category>
            <enclosure url="https://therouter.ai/news/claude-code-2273-gateway-hint-headers-classifier-flip-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4 Pro Survives Its Own Deadline: What the Reversal Means for Your Routing Policy]]></title>
            <link>https://therouter.ai/news/deepseek-v4-pro-survives-september-14-routing-decision/</link>
            <guid>https://therouter.ai/news/deepseek-v4-pro-survives-september-14-routing-decision/</guid>
            <pubDate>Mon, 14 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[DeepSeek announced on September 10 that V4 Pro would be retired today at 04:00 UTC. Instead, they reversed course in response to user demand, keeping V4 Pro live at unchanged pricing. Here is what the two-model landscape now looks like and which workloads belong on which.]]></description>
            <category>deepseek</category>
            <category>llm-api-ops</category>
            <category>operator</category>
            <category>model-routing</category>
            <category>pricing</category>
            <enclosure url="https://therouter.ai/news/deepseek-v4-pro-survives-september-14-routing-decision/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI Now Enforces API Key Expiration at the Org Level: What Gateway Operators Must Audit Today]]></title>
            <link>https://therouter.ai/news/openai-api-key-expiration-enforcement-operator/</link>
            <guid>https://therouter.ai/news/openai-api-key-expiration-enforcement-operator/</guid>
            <pubDate>Sun, 13 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI now lets admins enforce a maximum key lifetime at the org or project level. Existing keys are not retroactively shortened, but every new key created after the policy must expire within the limit — your gateway is the riskiest place to hold a long-lived key.]]></description>
            <category>openai</category>
            <category>governance</category>
            <category>billing</category>
            <category>operator</category>
            <enclosure url="https://therouter.ai/news/openai-api-key-expiration-enforcement-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI Prompt Cache Diagnostics Is GA: What Gateway Operators Must Not Strip]]></title>
            <link>https://therouter.ai/news/openai-prompt-cache-diagnostics-ga-routing-operator/</link>
            <guid>https://therouter.ai/news/openai-prompt-cache-diagnostics-ga-routing-operator/</guid>
            <pubDate>Sun, 13 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI's Prompt Cache Diagnostics tool is now GA in the Responses API. It adds comparison_response_id and a reason field that explain cache misses—and breaks silently if your AI gateway strips prompt_cache_options.]]></description>
            <category>openai</category>
            <category>prompt-caching</category>
            <category>api</category>
            <category>routing</category>
            <category>cost</category>
            <enclosure url="https://therouter.ai/news/openai-prompt-cache-diagnostics-ga-routing-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.269: Three Operator Changes Hidden Inside a 60-Fix Release]]></title>
            <link>https://therouter.ai/news/claude-code-2269-gateway-timeout-concurrent-agents-deny-rule-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2269-gateway-timeout-concurrent-agents-deny-rule-operator/</guid>
            <pubDate>Sat, 12 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.269 ships with a gateway discovery timeout override, a hard cap on concurrent workflow agents, and a fix for deny rules that were silently applying beyond their config source. Each one changes how operators govern Claude Code at scale.]]></description>
            <category>claude-code</category>
            <category>operator</category>
            <category>gateway</category>
            <category>coding-agents</category>
            <category>security</category>
            <category>prompt-cache</category>
            <enclosure url="https://therouter.ai/news/claude-code-2269-gateway-timeout-concurrent-agents-deny-rule-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI Enforces API Key Expiration at the Org Level: What Every Gateway Operator Must Audit Now]]></title>
            <link>https://therouter.ai/news/openai-api-key-expiration-org-enforcement/</link>
            <guid>https://therouter.ai/news/openai-api-key-expiration-org-enforcement/</guid>
            <pubDate>Sat, 12 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI now lets org admins enforce a maximum key lifetime across all projects. Combined with Anthropic's mandatory expiration policy from July, both major providers now require key rotation — with meaningfully different enforcement models operators must handle separately.]]></description>
            <category>openai</category>
            <category>billing-governance</category>
            <category>operator</category>
            <category>api-keys</category>
            <category>security</category>
            <enclosure url="https://therouter.ai/news/openai-api-key-expiration-org-enforcement/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.268: The Three-Version Gateway Breakage Operators Missed, and Four New Controls]]></title>
            <link>https://therouter.ai/news/claude-code-2268-anthropic-base-url-gateway-operator-fix/</link>
            <guid>https://therouter.ai/news/claude-code-2268-anthropic-base-url-gateway-operator-fix/</guid>
            <pubDate>Fri, 11 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Since 2.1.265, every turn through a third-party Anthropic-compatible endpoint failed with HTTP 400. Version 2.1.268 fixes it and adds gatewayInternalNetworks CIDR policy, pricing sync to gateway.yaml, and byte-stable prompt caches for Bedrock, Vertex, and Foundry.]]></description>
            <category>claude-code</category>
            <category>gateway</category>
            <category>operator</category>
            <category>coding-agents</category>
            <category>bedrock</category>
            <category>vertex</category>
            <category>prompt-cache</category>
            <enclosure url="https://therouter.ai/news/claude-code-2268-anthropic-base-url-gateway-operator-fix/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI Agents API Beta: The New Gateway Bypass Operators Need to Account For]]></title>
            <link>https://therouter.ai/news/openai-agents-api-beta-gateway-bypass-operator/</link>
            <guid>https://therouter.ai/news/openai-agents-api-beta-gateway-bypass-operator/</guid>
            <pubDate>Fri, 11 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI's Agents API beta ships a dedicated session harness under client.beta.agents that doesn't route through /v1/chat/completions. For teams using AI gateways, this means billing blind spots, missing audit trails, and a new API key scope to manage.]]></description>
            <category>openai</category>
            <category>agents</category>
            <category>gateway</category>
            <category>operator</category>
            <category>billing</category>
            <category>routing</category>
            <category>api-design</category>
            <enclosure url="https://therouter.ai/news/openai-agents-api-beta-gateway-bypass-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.267: Operator Effort Caps, Prompt-Cache Stability, and a Marketplace Containment Fix]]></title>
            <link>https://therouter.ai/news/claude-code-2-1-267-maxeffortlevel-prompt-cache-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2-1-267-maxeffortlevel-prompt-cache-operator/</guid>
            <pubDate>Thu, 10 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Three operator-facing changes land in 2.1.267: a new maxEffortLevel cap across Bedrock, Vertex, and Foundry; a flag to stop prompt-snapshot reuse during iteration; and a marketplace containment bypass fix. Plus twelve prompt-cache stability improvements.]]></description>
            <category>claude-code</category>
            <category>operator</category>
            <category>prompt-cache</category>
            <category>security</category>
            <category>gateway</category>
            <category>bedrock</category>
            <category>vertex</category>
            <enclosure url="https://therouter.ai/news/claude-code-2-1-267-maxeffortlevel-prompt-cache-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4.1-Flash Lands With Native Vision — and Kills deepseek-v4-pro in Four Days]]></title>
            <link>https://therouter.ai/news/deepseek-v4-1-flash-multimodal-v4-pro-routing-change/</link>
            <guid>https://therouter.ai/news/deepseek-v4-1-flash-multimodal-v4-pro-routing-change/</guid>
            <pubDate>Thu, 10 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[DeepSeek V4.1-Flash ships today under a new API name with native multimodal. On September 14, every deepseek-v4-pro request silently reroutes to V4.1-Flash at Flash pricing. Here is what the capability delta means and what to pin before the deadline.]]></description>
            <category>deepseek</category>
            <category>llm-api-ops</category>
            <category>operator</category>
            <category>multimodal</category>
            <category>model-routing</category>
            <category>pricing</category>
            <enclosure url="https://therouter.ai/news/deepseek-v4-1-flash-multimodal-v4-pro-routing-change/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.266 Fixes a Gateway-Breakage Regression: Who Is Affected and What to Do]]></title>
            <link>https://therouter.ai/news/claude-code-2266-gateway-use-gateway-regression-fix/</link>
            <guid>https://therouter.ai/news/claude-code-2266-gateway-use-gateway-regression-fix/</guid>
            <pubDate>Wed, 09 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[2.1.265 silently broke all gateway and proxy setups that had CLAUDE_CODE_USE_GATEWAY set alongside an API key or custom auth. Here's the exact affected configuration, the error, and what the release adds for gateway operators.]]></description>
            <category>claude-code</category>
            <category>gateway</category>
            <category>operator</category>
            <category>reliability</category>
            <enclosure url="https://therouter.ai/news/claude-code-2266-gateway-use-gateway-regression-fix/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Fable 5.1 Per-Message Effort and Mid-Session Tool Changes: Three Beta Controls That Don't Break Your Cache]]></title>
            <link>https://therouter.ai/news/claude-fable-5-1-per-message-effort-tool-changes-operator/</link>
            <guid>https://therouter.ai/news/claude-fable-5-1-per-message-effort-tool-changes-operator/</guid>
            <pubDate>Wed, 09 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Anthropic's September 1 API release added three beta controls — per-message effort, turn-scoped system messages, and mid-conversation tool changes — that let operators adjust compute intensity and tool sets in long agentic sessions without invalidating the cache prefix.]]></description>
            <category>claude-api</category>
            <category>prompt-caching</category>
            <category>llm-api-ops</category>
            <category>agentic-systems</category>
            <enclosure url="https://therouter.ai/news/claude-fable-5-1-per-message-effort-tool-changes-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[ant apply Brings GitOps to Claude Agents: What Every Gateway Operator Must Track]]></title>
            <link>https://therouter.ai/news/ant-apply-claude-agent-gitops-operator/</link>
            <guid>https://therouter.ai/news/ant-apply-claude-agent-gitops-operator/</guid>
            <pubDate>Tue, 08 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Anthropic's ant CLI v1.30.0 ships ant apply: declare agents, environments, and deployments as repo files. For gateway operators, the model field now governs routing, and claude-lock.json tracks drift.]]></description>
            <category>claude-api</category>
            <category>agent-ops</category>
            <category>governance</category>
            <category>operator-workflow</category>
            <enclosure url="https://therouter.ai/news/ant-apply-claude-agent-gitops-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Fable 5.1's Cache Read Cut and Effort Parameter Are the Two Changes Every Gateway Operator Must Price In]]></title>
            <link>https://therouter.ai/news/claude-fable-5-1-effort-cache-read-pricing-operator/</link>
            <guid>https://therouter.ai/news/claude-fable-5-1-effort-cache-read-pricing-operator/</guid>
            <pubDate>Tue, 08 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Cache reads on claude-fable-5-1 drop to $0.25/MTok — 4x cheaper than other Claude models — and output_config.effort lets operators trade thinking depth for token spend per request. What both changes mean for routing, billing, and upgrade decisions.]]></description>
            <category>anthropic</category>
            <category>model-pricing</category>
            <category>routing</category>
            <category>llm-api-ops</category>
            <enclosure url="https://therouter.ai/news/claude-fable-5-1-effort-cache-read-pricing-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.260: The Fable 5.1 Prompt-Cache Bug That Cost Every Tool-Using Agent]]></title>
            <link>https://therouter.ai/news/claude-code-2260-fable-prompt-cache-tool-context-fix-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2260-fable-prompt-cache-tool-context-fix-operator/</guid>
            <pubDate>Mon, 07 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Fable 5.1 prompt caching silently excluded post-tool-result context, re-sending it uncached on every tool-call turn. Claude Code 2.1.260 fixes it — along with cache-miss diagnostics, a Read() permission revert, and a managed CLAUDE.md governance change.]]></description>
            <category>claude-code</category>
            <category>fable-5-1</category>
            <category>prompt-cache</category>
            <category>gateway</category>
            <category>operator</category>
            <enclosure url="https://therouter.ai/news/claude-code-2260-fable-prompt-cache-tool-context-fix-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI Splits the Rate-Limit 429: What 'slow_down' vs 'server_is_overloaded' Mean for Your Retry Policy]]></title>
            <link>https://therouter.ai/news/openai-429-slow-down-503-overloaded-retry-routing/</link>
            <guid>https://therouter.ai/news/openai-429-slow-down-503-overloaded-retry-routing/</guid>
            <pubDate>Mon, 07 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI now returns distinct error codes for two previously identical 429s — traffic that ramps too fast gets slow_down, model saturation gets a 503. Your retry logic must treat them differently, and most AI gateway clients currently don't.]]></description>
            <category>openai</category>
            <category>rate-limits</category>
            <category>retry</category>
            <category>gateway</category>
            <category>operator</category>
            <enclosure url="https://therouter.ai/news/openai-429-slow-down-503-overloaded-retry-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.261 Gateway Policy Telemetry Operator Changes: What Proxies Must Fix]]></title>
            <link>https://therouter.ai/news/claude-code-2-1-261-gateway-policy-telemetry-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2-1-261-gateway-policy-telemetry-operator/</guid>
            <pubDate>Sat, 05 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.261 gateway policy telemetry operator changes expose four proxy failure modes: hidden organization policy fetches, malformed client IP parsing, JSON/protobuf OTEL drift, and Bedrock setup checks behind TLS inspection.]]></description>
            <category>claude-code</category>
            <category>gateway</category>
            <category>telemetry</category>
            <category>operator</category>
            <category>proxy</category>
            <enclosure url="https://therouter.ai/news/claude-code-2-1-261-gateway-policy-telemetry-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[GPT-6 Astra Cuts Chat Completions Tool Calling: What Operators Must Change Before Routing Traffic]]></title>
            <link>https://therouter.ai/news/gpt-6-astra-tool-calling-responses-api-operator-migration/</link>
            <guid>https://therouter.ai/news/gpt-6-astra-tool-calling-responses-api-operator-migration/</guid>
            <pubDate>Fri, 04 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[GPT-6 Astra refuses Chat Completions tool calls, drops temperature and top_p, adds two new error codes operators must handle, and runs async misalignment monitoring that changes the latency contract for Responses API users.]]></description>
            <category>openai</category>
            <category>gpt-6-astra</category>
            <category>responses-api</category>
            <category>tool-calling</category>
            <category>operator</category>
            <category>migration</category>
            <enclosure url="https://therouter.ai/news/gpt-6-astra-tool-calling-responses-api-operator-migration/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.259: Org-Level MCP Server Push and the Concurrent Session State Fix Operators Missed]]></title>
            <link>https://therouter.ai/news/claude-code-2259-managed-mcp-servers-concurrent-session-fix/</link>
            <guid>https://therouter.ai/news/claude-code-2259-managed-mcp-servers-concurrent-session-fix/</guid>
            <pubDate>Thu, 03 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.259 ships managedMcpServers for org-wide HTTP/SSE MCP rollout, breaks the allowedMcpServers filter for managed entries, and fixes a concurrent-session bug that was silently corrupting workspace trust and MCP state in multi-agent CI runs.]]></description>
            <category>claude-code</category>
            <category>mcp</category>
            <category>operator</category>
            <category>ci</category>
            <category>governance</category>
            <enclosure url="https://therouter.ai/news/claude-code-2259-managed-mcp-servers-concurrent-session-fix/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Fable 5.1 Breaks Two Operator Assumptions: `tool_choice` Force-Calls and Thinking Block Reuse]]></title>
            <link>https://therouter.ai/news/claude-fable-5-1-tool-choice-400-thinking-block-binding-operator/</link>
            <guid>https://therouter.ai/news/claude-fable-5-1-tool-choice-400-thinking-block-binding-operator/</guid>
            <pubDate>Thu, 03 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Fable 5.1 breaks two patterns operators rely on: `tool_choice: any` and `tool_choice: tool` now return 400, and thinking blocks are version-gated with prefix-mismatch enforcement for new accounts. Migrate both before routing production traffic.]]></description>
            <category>anthropic</category>
            <category>claude</category>
            <category>fable-5-1</category>
            <category>tool-use</category>
            <category>api-breaking-change</category>
            <category>thinking</category>
            <category>operator</category>
            <category>llm-api-ops</category>
            <enclosure url="https://therouter.ai/news/claude-fable-5-1-tool-choice-400-thinking-block-binding-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Fable 5.1 Arrives: Containment Escape Blocks Cloud Credential Fetches by Default]]></title>
            <link>https://therouter.ai/news/claude-fable-5-1-containment-escape-operator-routing/</link>
            <guid>https://therouter.ai/news/claude-fable-5-1-containment-escape-operator-routing/</guid>
            <pubDate>Wed, 02 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code ships Fable 5.1 with 1M context and a hard auto-mode gate on cloud metadata fetches — here is what every operator running production pipelines must audit now.]]></description>
            <category>claude-code</category>
            <category>model-releases</category>
            <category>coding-agents</category>
            <category>governance</category>
            <enclosure url="https://therouter.ai/news/claude-fable-5-1-containment-escape-operator-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI mTLS and X.509 Workload Identity Are Now GA: What Every Gateway Operator Must Audit]]></title>
            <link>https://therouter.ai/news/openai-mtls-x509-gateway-routing-operator/</link>
            <guid>https://therouter.ai/news/openai-mtls-x509-gateway-routing-operator/</guid>
            <pubDate>Tue, 01 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI mutual TLS and X.509 workload identity federation went GA on August 29. Once an organization activates mTLS, every proxy or gateway forwarding its API traffic must present a valid client certificate or requests fail at the TLS handshake.]]></description>
            <category>openai</category>
            <category>security</category>
            <category>gateway</category>
            <category>authentication</category>
            <enclosure url="https://therouter.ai/news/openai-mtls-x509-gateway-routing-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.251: The PreModelSwitch Hook Is Now Your Model Governance Control Plane]]></title>
            <link>https://therouter.ai/news/claude-code-2251-premodelswitch-hook-spend-limit-gateway/</link>
            <guid>https://therouter.ai/news/claude-code-2251-premodelswitch-hook-spend-limit-gateway/</guid>
            <pubDate>Mon, 31 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.251 adds PreModelSwitch and PostModelSwitch hooks that let operators block, confirm, or annotate model switches in real time. Combined with new spend-limit visibility and per-session cache metrics, this release makes the client itself a usable governance surface.]]></description>
            <category>claude-code</category>
            <category>coding-agents</category>
            <category>gateway-operations</category>
            <category>governance</category>
            <enclosure url="https://therouter.ai/news/claude-code-2251-premodelswitch-hook-spend-limit-gateway/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Anthropic Model Hardware Standard: The New Safety Boundary for Physical AI Agents]]></title>
            <link>https://therouter.ai/news/anthropic-mhs-physical-agent-safety-routing/</link>
            <guid>https://therouter.ai/news/anthropic-mhs-physical-agent-safety-routing/</guid>
            <pubDate>Fri, 28 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Anthropic Model Hardware Standard turns lab devices into discoverable agent tools. For operators, the critical work is routing authority, safety limits, and audit paths before code touches hardware.]]></description>
            <category>anthropic</category>
            <category>agents</category>
            <category>mcp</category>
            <category>safety</category>
            <category>routing</category>
            <enclosure url="https://therouter.ai/news/anthropic-mhs-physical-agent-safety-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.247: The /claude-api cost-optimize Command Changes How Operators Audit API Spend]]></title>
            <link>https://therouter.ai/news/claude-code-2247-cost-optimize-subagent-fallback-operator/</link>
            <guid>https://therouter.ai/news/claude-code-2247-cost-optimize-subagent-fallback-operator/</guid>
            <pubDate>Thu, 27 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.247 ships a built-in cost-profiling command that walks operators through caching, batch, token hygiene, and model-choice levers, and fixes a silent failure where sub-agents died on their first model call without activating the gateway fallback chain.]]></description>
            <category>claude-code</category>
            <category>coding-agents</category>
            <category>cost-optimization</category>
            <category>gateway-operations</category>
            <enclosure url="https://therouter.ai/news/claude-code-2247-cost-optimize-subagent-fallback-operator/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4-Flash-Vision-Exp Adds Vision to the DeepSeek API: What Your Routing Layer Must Handle Now]]></title>
            <link>https://therouter.ai/news/deepseek-v4-flash-vision-exp-multimodal-routing/</link>
            <guid>https://therouter.ai/news/deepseek-v4-flash-vision-exp-multimodal-routing/</guid>
            <pubDate>Thu, 27 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[DeepSeek launched its first multimodal model on August 21. The critical routing problem is not the model itself — it's that vision requests require a content array where your proxy may only be passing a string.]]></description>
            <category>deepseek</category>
            <category>multimodal</category>
            <category>vision</category>
            <category>routing-policy</category>
            <category>llm-api-ops</category>
            <category>model-releases</category>
            <enclosure url="https://therouter.ai/news/deepseek-v4-flash-vision-exp-multimodal-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[gpt-image-2 Now Ships Native Alpha Channels: What Your Image Pipeline Needs to Change]]></title>
            <link>https://therouter.ai/news/openai-gpt-image-2-transparent-background-api-routing/</link>
            <guid>https://therouter.ai/news/openai-gpt-image-2-transparent-background-api-routing/</guid>
            <pubDate>Wed, 26 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI added transparent background support to gpt-image-2 on August 20 — in preview across both the Images API and the Responses API image generation tool. One new parameter, one silent failure mode on jpeg, and a preview-status caveat that matters if you have ZDR.]]></description>
            <category>openai</category>
            <category>image-generation</category>
            <category>routing</category>
            <category>api</category>
            <enclosure url="https://therouter.ai/news/openai-gpt-image-2-transparent-background-api-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI's Per-Request Regional Routing: One Key, Any Region, No New Projects]]></title>
            <link>https://therouter.ai/news/openai-per-request-regional-endpoint-routing/</link>
            <guid>https://therouter.ai/news/openai-per-request-regional-endpoint-routing/</guid>
            <pubDate>Wed, 26 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI now lets API operators select a processing region per request using a prefixed base URL with one Global project key. Multi-region gateway architecture changes: one credential fans out to us.api.openai.com or eu.api.openai.com based on per-request routing logic.]]></description>
            <category>openai</category>
            <category>data-residency</category>
            <category>routing</category>
            <category>enterprise</category>
            <enclosure url="https://therouter.ai/news/openai-per-request-regional-endpoint-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.243: Three Managed Settings That Reshape How Operators Control Models, Cache, and Cost]]></title>
            <link>https://therouter.ai/news/claude-code-2243-model-picker-cache-ttl-cost-attribution/</link>
            <guid>https://therouter.ai/news/claude-code-2243-model-picker-cache-ttl-cost-attribution/</guid>
            <pubDate>Tue, 25 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.243 adds modelPicker, promptCacheTtl, and modelPricing — three managed settings that let gateway operators lock the model menu, tune cache TTLs per agent tier, and replace list-price cost figures with contracted rates.]]></description>
            <category>claude-code</category>
            <category>coding-agents</category>
            <category>operator-settings</category>
            <category>prompt-cache</category>
            <category>cost-attribution</category>
            <enclosure url="https://therouter.ai/news/claude-code-2243-model-picker-cache-ttl-cost-attribution/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code Data Residency Workspaces Now Carry a 1.1× Cost Premium Operators Must Route Around]]></title>
            <link>https://therouter.ai/news/claude-code-2245-data-residency-workspace-cost-routing/</link>
            <guid>https://therouter.ai/news/claude-code-2245-data-residency-workspace-cost-routing/</guid>
            <pubDate>Tue, 25 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.239 started surfacing the 1.1× US-only-inference premium in /cost and status-line estimates for data-residency workspaces. Teams routing through Bedrock or Vertex avoid the surcharge by design — here is what that means for your provider selection.]]></description>
            <category>claude-code</category>
            <category>coding-agents</category>
            <category>billing</category>
            <category>data-residency</category>
            <category>operator</category>
            <enclosure url="https://therouter.ai/news/claude-code-2245-data-residency-workspace-cost-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.239: The Proxy Bug That Silently Doubled Your Bedrock API Bill]]></title>
            <link>https://therouter.ai/news/claude-code-2239-bedrock-proxy-content-type-billing-fix/</link>
            <guid>https://therouter.ai/news/claude-code-2239-bedrock-proxy-content-type-billing-fix/</guid>
            <pubDate>Mon, 24 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[A stripped Content-Type response header from proxies, API gateways, or load balancers caused Claude Code to re-run every Bedrock turn in non-streaming mode — billing you twice per request. Here is how to verify exposure and what else 2.1.239 fixes for gateway operators.]]></description>
            <category>coding-agents</category>
            <category>claude-code</category>
            <category>bedrock</category>
            <category>proxy</category>
            <category>billing</category>
            <enclosure url="https://therouter.ai/news/claude-code-2239-bedrock-proxy-content-type-billing-fix/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[qwen3.8-max DashScope Routing Policy: Endpoint, Reasoning, and Region Checks]]></title>
            <link>https://therouter.ai/news/qwen3-8-max-dashscope-routing-policy/</link>
            <guid>https://therouter.ai/news/qwen3-8-max-dashscope-routing-policy/</guid>
            <pubDate>Sun, 23 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[qwen3.8-max DashScope routing policy now starts with region-scoped endpoints, Responses API reasoning, and whether your gateway preserves reasoning_content.]]></description>
            <category>qwen</category>
            <category>dashscope</category>
            <category>routing</category>
            <category>llm-api-ops</category>
            <enclosure url="https://therouter.ai/news/qwen3-8-max-dashscope-routing-policy/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.238 Gives Self-Hosted Runners Dynamic Proxy Authorization and Graceful Shutdown]]></title>
            <link>https://therouter.ai/news/claude-code-2238-proxy-auth-runner-graceful-shutdown/</link>
            <guid>https://therouter.ai/news/claude-code-2238-proxy-auth-runner-graceful-shutdown/</guid>
            <pubDate>Fri, 21 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[2.1.238 adds --proxy-authorization-command and --proxy-authorization-file to self-hosted runners, ending static-header workarounds for corporate proxies. It also introduces --defer-shutdown-max-min and extends headersHelper to marketplace catalog fetches.]]></description>
            <category>claude-code</category>
            <category>self-hosted-runner</category>
            <category>operator</category>
            <category>proxy</category>
            <category>enterprise</category>
            <enclosure url="https://therouter.ai/news/claude-code-2238-proxy-auth-runner-graceful-shutdown/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.237 Fixes Prompt Caching on Every LLM Gateway — and 2.1.236 Adds a Default Model You Can Override]]></title>
            <link>https://therouter.ai/news/claude-code-2237-gateway-prompt-cache-fix-default-model/</link>
            <guid>https://therouter.ai/news/claude-code-2237-gateway-prompt-cache-fix-default-model/</guid>
            <pubDate>Thu, 20 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.237 silently fixed a bug that invalidated prompt caching for every session routed through a custom base URL. If you've been routing Claude Code through a gateway, you've been paying full context-window cost on every turn since at least 2.1.229.]]></description>
            <category>claude-code</category>
            <category>gateway</category>
            <category>prompt-caching</category>
            <category>operator</category>
            <category>bedrock</category>
            <category>vertex</category>
            <enclosure url="https://therouter.ai/news/claude-code-2237-gateway-prompt-cache-fix-default-model/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Anthropic Inference Hooks Put a Pre-Inference Gate at the Provider Layer: What It Means for Your Routing Architecture]]></title>
            <link>https://therouter.ai/news/anthropic-inference-hooks-enterprise-gateway-architecture/</link>
            <guid>https://therouter.ai/news/anthropic-inference-hooks-enterprise-gateway-architecture/</guid>
            <pubDate>Wed, 19 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Anthropic's new Inference Hooks let enterprise organizations intercept every governed Claude prompt before the model runs. For teams already filtering at the gateway layer, this creates a dual-gate architecture that changes where enforcement belongs.]]></description>
            <category>anthropic</category>
            <category>enterprise</category>
            <category>security</category>
            <category>governance</category>
            <category>routing</category>
            <enclosure url="https://therouter.ai/news/anthropic-inference-hooks-enterprise-gateway-architecture/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Text Watermarking Is Coming to All Operators: What the SynthID-Text Detection API Means for Your Pipeline]]></title>
            <link>https://therouter.ai/news/anthropic-claude-text-watermark-operator-pipeline/</link>
            <guid>https://therouter.ai/news/anthropic-claude-text-watermark-operator-pipeline/</guid>
            <pubDate>Tue, 18 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Future Claude models will embed a SynthID-Text watermark globally, with no opt-out. A detection API is coming. Operators running content-review or compliance pipelines need to understand what the watermark proves — and what pipeline modifications can silently degrade it.]]></description>
            <category>anthropic</category>
            <category>claude</category>
            <category>watermark</category>
            <category>compliance</category>
            <category>eu-ai-act</category>
            <category>content-provenance</category>
            <category>operator</category>
            <category>governance</category>
            <enclosure url="https://therouter.ai/news/anthropic-claude-text-watermark-operator-pipeline/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4 Reasoning Effort Lands in Three API Formats: Why Your Proxy Is Probably Getting It Wrong]]></title>
            <link>https://therouter.ai/news/deepseek-v4-reasoning-effort-proxy-routing/</link>
            <guid>https://therouter.ai/news/deepseek-v4-reasoning-effort-proxy-routing/</guid>
            <pubDate>Tue, 18 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[DeepSeek V4-Pro and V4-Flash now expose three-tier reasoning effort control across OpenAI, Anthropic, and Responses API formats. The catch: each format uses a different field, and proxies that strip extra_body silently override your effort to max.]]></description>
            <category>deepseek</category>
            <category>api</category>
            <category>routing</category>
            <category>llm</category>
            <enclosure url="https://therouter.ai/news/deepseek-v4-reasoning-effort-proxy-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.233: Three Gateway Changes Every Operator Must Know]]></title>
            <link>https://therouter.ai/news/claude-code-2233-user-identity-gateway-spend-attribution/</link>
            <guid>https://therouter.ai/news/claude-code-2233-user-identity-gateway-spend-attribution/</guid>
            <pubDate>Mon, 17 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Version 2.1.233 adds forward_user_identity to surface per-developer identity through the apps gateway to any downstream proxy, fixes silent 400/413 swallowing on Vertex, Foundry, and AWS, and closes an infinite-reconnect loop in MCP v2 on serverless hosts.]]></description>
            <category>claude-code</category>
            <category>gateway</category>
            <category>operator</category>
            <category>spend-attribution</category>
            <category>mcp</category>
            <enclosure url="https://therouter.ai/news/claude-code-2233-user-identity-gateway-spend-attribution/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4-Pro-0813 and the August 16 Pricing Cliff: What Every Gateway Operator Must Do in the Next 48 Hours]]></title>
            <link>https://therouter.ai/news/deepseek-v4-pro-august16-pricing-agent-routing/</link>
            <guid>https://therouter.ai/news/deepseek-v4-pro-august16-pricing-agent-routing/</guid>
            <pubDate>Fri, 14 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[DeepSeek's V4-Pro alias just upgraded to a new GA model with top agent benchmarks, and peak/off-peak pricing goes live August 16. The cost math for production routing teams has fundamentally changed.]]></description>
            <category>deepseek</category>
            <category>api-pricing</category>
            <category>model-routing</category>
            <category>agent-routing</category>
            <enclosure url="https://therouter.ai/news/deepseek-v4-pro-august16-pricing-agent-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[OpenAI Ultrafast Mode Routing: GPT-5.6 Sol Gets a 14x Preview Tier]]></title>
            <link>https://therouter.ai/news/openai-ultrafast-mode-gpt56-sol-routing/</link>
            <guid>https://therouter.ai/news/openai-ultrafast-mode-gpt56-sol-routing/</guid>
            <pubDate>Fri, 14 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI Ultrafast mode routing adds a limited-preview GPT-5.6 Sol tier that is up to 14x faster than Standard, forcing operators to split latency-critical traffic from cost-sensitive fallback lanes.]]></description>
            <category>openai</category>
            <category>model-routing</category>
            <category>latency</category>
            <category>api-ops</category>
            <enclosure url="https://therouter.ai/news/openai-ultrafast-mode-gpt56-sol-routing/cover.webp" length="0" type="image/webp"/>
        </item>
        <item>
            <title><![CDATA[Claude Code 2.1.229 Adds SSE Keepalives for Vertex and Bedrock — And Locks Down the Git Push Path]]></title>
            <link>https://therouter.ai/news/claude-code-2229-sse-keepalive-vertex-bedrock-gateway-streaming/</link>
            <guid>https://therouter.ai/news/claude-code-2229-sse-keepalive-vertex-bedrock-gateway-streaming/</guid>
            <pubDate>Thu, 13 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Claude Code 2.1.229 injects SSE keepalive pings into gateway streaming responses during long thinking pauses, preventing silent idle-timeout disconnects on Vertex AI and Bedrock. Plus: /commit-push-pr now blocks dangerous git flags and the sandbox enforces IPv6 fail-closed.]]></description>
            <category>claude-code</category>
            <category>streaming</category>
            <category>vertex-ai</category>
            <category>bedrock</category>
            <category>gateway</category>
            <category>operator</category>
            <category>security</category>
            <enclosure url="https://therouter.ai/news/claude-code-2229-sse-keepalive-vertex-bedrock-gateway-streaming/cover.webp" length="0" type="image/webp"/>
        </item>
    </channel>
</rss>