OpenRouter
1 Tool · 213 Intel
OpenRouter is a unified API platform, founded in early 2023, that gives developers access to hundreds of large language models from OpenAI, Anthropic, Google, and dozens of other providers through a single endpoint. It handles routing, caching, and uptime fallback across providers and charges a small fee on top of underlying model costs rather than marking up prices. The company raised a $12.5 million seed round in early 2025 followed by a $28 million Series A led by Menlo Ventures.
Is this you? Sign in with X to claim this profile.
Tools
Async Python agent toolkit for OpenRouter with tools, streaming, and human approval gates.
Intel
OpenRouter is joining Stripe. The company says the name, product, roadmap and multi-model commitment continue as before. It describes itself as the largest AI marketplace and gateway, routing more than 10 trillion tokens daily across 400+ models for 10m+ developers and businesses.
OpenRouter names its stealth model: Ox Alpha was GLM-5.3-Flash, which processed over 20 trillion tokens in six days. Hybrid sparse and linear attention keeps its 1M context usable, output caps at 131K, max reasoning is the default, and launch pricing is half off through September 9.
OpenRouter data on GPT-5.6 discounts: Luna token volume rose 13.8x, Terra 5.6x, Sol 1.1x. Most growth was net new usage or share taken from rivals, and daily volume stayed higher after discounts ended. Based on the thread; the full report and charts were not read.
DeepSeek V4 Pro 0813 is live on OpenRouter. DeepSeek reports gains over V4 Pro Preview: DeepSWE 62.7, CyberGym 83.3, NL2Repo 61.5 and Terminal Bench 2.1 87.9. More providers are expected soon. Independent verification is not yet available.
OpenRouter discounted GPT-5.6 Sol by 50 percent through September 18, applied automatically on the openai/gpt-5.6-sol slug and covering batch, flex and priority tiers. Flex reached $1.25 in and $7.50 out. Limited to the OpenAI provider and non-BYOK requests.
OpenRouter serves GLM-5.3 and points at where the gains came from: the same base model as GLM-5.2, with post-training alone moving Terminal-Bench 3.0 from 4.6 to 28.3 and DeepSWE v1.1 from 46.2 to 66.9.
OpenRouter added a Muse Spark 1.2 contributor tier at $0.10 per million input and $0.20 per million output tokens. The discount depends on Meta using prompts and outputs to improve its products, which suits experimentation and early projects but not sensitive work.
OpenRouter launched Ori Pi, which runs the Pi coding agent on OpenRouter with one install and OAuth sign-in. The model picker honors workspace guardrails, /fast routes for speed by provider, and /zdr limits a session to zero-retention endpoints. Both settings are saved with the session.
OpenRouter announces Inception's Mercury 2.5 Preview, a reasoning LLM that generates tokens in parallel and is quoted at 1,107 tokens per second. It offers tunable reasoning, parallel tool calls and schema-aligned JSON, and is available exclusively on OpenRouter during the preview.
DeepSeek V4 Flash Vision Exp is available on OpenRouter with image input at standard V4 Flash pricing. OpenRouter reports parity with V4 Flash 0731 on text agent benchmarks and wins over Opus-4.8 on Agents' Last Exam and ZeroBench. Figures are unverified.
OpenRouter lists Gemini 3.8 Flash at 3.7 Flash's intro price, claiming gains on its shown benchmarks. A new video option, processing set to agentic, lets the model navigate the timeline itself rather than sample frames at a fixed rate, and it also works on earlier Flash models. Benchmarks were not i
OpenRouter launched Ori DeepSeek Harness, a one-command setup that connects DeepSeek Harness to the full OpenRouter catalog. It installs the harness if missing, stores the key in the credential store, refreshes models at boot, and adds /fast and /zdr session controls.
OpenRouter added a Cost per Session data layer: median real cost of tasks in each major harness over 30 days, which captures token efficiency that per-token pricing misses. Rankings are published and the data is queryable via a Dataset API under CC BY 4.0.
Meta's Muse Spark 1.2 is available on OpenRouter at $1.25 per million input and $4.25 per million output tokens, with global access widened across the Muse Spark line. It is pitched at multi-file refactors and long debugging sessions. Claims come from OpenRouter, not benchmarks.
OpenRouter now serves Claude Fable 5.1 at Fable 5 pricing ($10 input, $50 output per million tokens) with prompt cache reads down 75% to $0.25 per million. Stated gains target agentic coding, long-running workflows, visual code generation, finance and analysis. Capability claims come from the thread
NVIDIA's Nemotron 3.5 Lightning, a 30B hybrid MoE with 3B active parameters distilled from Nemotron 3 Ultra, is available on OpenRouter. NVIDIA claims up to 4x throughput and 30 percent faster task completion versus similar models. Figures are unverified vendor claims.
Thinking Machines now serves its open-weight Inkling (975B total, 41B active) and Inkling Small (276B/12B) on OpenRouter, free only inside agentic harnesses. Both take text, image and audio with 1M context. Prompts are logged for training; weights are Apache 2.0.
OpenRouter added MiniMax H3 Max, a video model it says finishes the same prompt in about 30 seconds instead of 3.5 minutes and costs $0.40 per clip instead of $0.65 for H3. The comparison is the provider's own and the demo video could not be read.
Meta's Muse Image is live on OpenRouter at $0.01 per image, grounded by search. It generates and edits complex images while leaving unrequested regions intact, a pricing and capability change for teams running high-volume image asset workflows.
Alibaba's Qwen3.8 Flash is now routable on OpenRouter. It is a multimodal reasoning model aimed at coding assistants and agentic workflows, with codebase and document analysis, desktop interaction, chart reading, and long video understanding.
Alibaba's Qwen3.8 Max is available on OpenRouter with 2.4T total and 95B active parameters, aimed at long-horizon coding and multimodal agents. Qwen reports large benchmark gains over its prior Max model, and open weights are promised within a week.
OpenRouter's Ori CLI now launches xAI's Grok Build in custom-endpoint mode with a run-scoped OpenRouter key. No browser login, Grok config untouched, model list drawn from your OpenRouter catalog, and xAI telemetry, error reporting and trace upload disabled.
OpenRouter launched video benchmarks that run video generation models on shared challenge prompts and compare quality, cost and generation time. Its thread says Seedance 2.5 leads on physics and prompt adherence, MiniMax H3 Max leads on speed and price, and custom prompts are supported.
OpenRouter's Agent SDK added lifecycle hooks in the callModel loop to block destructive commands, redact secrets, track cost, or rewrite tool arguments. Async tools add background and deferred modes, so a webhook or human review can resume a run later, with a task tool for status and cancel.