OpenRouter

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 measured what a heavy discount did to GPT-5.6 demand: Luna volume up 13.8x, Terra 5.6x, Sol 1.1x. Some traffic came from older OpenAI models, but most was net new usage plus share taken from competitors, and daily volume stayed higher once the discount ended.

DeepSeek V4 Pro 0813 is live with first-party gains over the V4 Pro Preview: DeepSWE 62.7, CyberGym 83.3, NL2Repo 61.5, Terminal Bench 2.1 87.9. Independent scores are not published yet, so treat the deltas as vendor claims.

GPT-5.6 Sol is discounted 50 percent on OpenRouter through September 18, applied automatically on the openai/gpt-5.6-sol slug and extending to batch, flex, and priority tiers. Flex reaches $1.25 in and $7.50 out. OpenAI provider only, 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.

Meta's Muse Spark 1.2 contributor tier is on OpenRouter at $0.10 per million input and $0.20 per million output tokens. The discount is paid for in data: prompts and outputs may be used to improve Meta products, which rules it out for sensitive work.

OpenRouter packages the Pi agent with its credentials and catalog: the model picker respects workspace guardrails, /fast routes to Anthropic's fast tier, OpenAI priority, or the highest-throughput provider, and /zdr restricts a session to zero-retention endpoints. Both toggles survive resume.

DeepSeek's V4 Flash Vision Exp accepts image input at standard V4 Flash pricing, holds parity with V4 Flash 0731 on text agent benchmarks, and is reported ahead of Opus-4.8 on Agents' Last Exam and ZeroBench.

Ori DeepSeek wires DeepSeek Harness to OpenRouter in one command: it installs the harness if missing, stores the key in the harness credential store, refreshes the model catalog on every boot, and adds /fast routing plus /zdr zero-retention sessions.

OpenRouter added cost per session: the median real cost of a task in each major harness over 30 days, which folds in the token efficiency that per-token pricing hides. Rankings are published and the underlying data is queryable under CC BY 4.0.

Meta's Muse Spark 1.2 is on OpenRouter at $1.25 per million input tokens and $4.25 per million output, with global access widened across the Muse Spark line. It targets multi-file refactors and tasks that run past a single prompt.

Nemotron 3.5 Lightning is a 30B hybrid MoE with 3B active parameters, distilled from Nemotron 3 Ultra and now on OpenRouter. NVIDIA claims up to 4x higher throughput and up to 30 percent faster task completion than comparable models in its class.

Thinking Machines serves Inkling (975B total, 41B active) and Inkling Small (276B/12B) on OpenRouter, free inside agentic harnesses. Open weights are Apache 2.0, context runs to 1M with text, image and audio input, and prompts are logged for training.

Muse Image from Meta is available on OpenRouter at $0.01 per image, grounded by search. It creates and edits complex images while preserving the regions you did not ask it to change, the pattern high-volume asset work needs.

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 live with 2.4T total parameters and 95B active, built for long-horizon coding and multimodal agent work. Qwen reports gains over 3.7 Max on PaperBench (64.8 to 93.0), OSWorld-Verified (73.3 to 86.1) and Vision2Web (42.1 to 69.0). Open weights are due next week.

ori grok starts xAI's Grok Build in custom-endpoint mode and hands it an OpenRouter key scoped to that run, so no browser login happens and your Grok config is unchanged. Model list becomes your OpenRouter catalog, private endpoints included, with telemetry and trace upload pinned off.

The OpenRouter Agent SDK gained lifecycle hooks in the callModel loop for blocking destructive commands, redacting secrets, metering cost, or rewriting tool arguments before execution. Async tools add background and deferred modes, so a webhook or human review can resume a run later.

BytePlus's Dreamina Seedance 2.5 is on OpenRouter as a joint audio-video model: up to 30 seconds per pass, extendable across rounds, with up to 50 references (30 images, 10 videos, 10 audio) and editing that can replace subjects or add and remove objects.

OpenRouter's task-level rankings, measured by real spend share rather than benchmark scores, now show DeepSeek V4 Pro leading shell execution and Kimi K3 leading tool dispatch. Both are open-weight models.

OpenRouter published a visual benchmark grid for image models: each model's output on hard prompts such as composing several inputs into one scene, showing a specific number of fingers, or removing the second cup in a row, sortable by price and generation time.

OpenRouter's New Defaults episode with Deedy Das of Menlo Ventures on inference economics: token prices down 40 percent while usage grew fourfold, why optimizing cost too early backfires, and why serious AI products end up spanning multiple models.

ori prime runs Prime Intellect's Agent against 500+ OpenRouter models without writing auth.json or models.json; the key arrives via a bundled extension scoped to the run. Ori strips leftover ANTHROPIC_API_KEY and OPENAI_API_KEY values, and /speed and /zdr persist with the session.

Alibaba's Wan 3.0 generates 2 to 30 second videos from text, images, or references at 480p, 720p, or 1080p, priced at $0.05, $0.10, and $0.20 per second with a limited-time 15 percent discount.

Black Forest Labs' FLUX 3 is live on OpenRouter's video API: one jointly trained family spanning video, audio, image, and action prediction, with text-to-video, multi-keyframe image-to-video, video continuation, and native audio including lipsync and ambience.

OpenRouter's Activity dashboard adds movement-ranked trends across models, keys, users and apps, an Explore builder for arbitrary metric and dimension combinations, and a matching Analytics API with /analytics/meta for discovery. An internal agent-run cost review found a $6.2K/month misrouted key.

Grok Imagine Image 2.0 is on OpenRouter from $0.04 per image, taking up to three reference photos per request. It plans typography, layout and multi-part visuals, and adds edits that change one element without disturbing the rest of the frame.

Muse Glimmer, the first open-weights release from Meta Superintelligence Labs, is a 30B dense text and image model under Apache 2.0, built as a reliable local agent, with reported scores of 75.5 on MCP Atlas and 51.2 on SWE-Bench Pro.

The :nitro and :floor variants no longer only sort existing endpoints. :nitro now admits priority service tier endpoints into the candidate pool, competing on measured throughput, and the ~latest alias resolves to a provider's newest model, so one slug stacks all three behaviors.

Black Forest Labs' FLUX Video Upscale is available through OpenRouter's video API, upscaling to 2K or 4K in two modes: precise, which is faster and holds identity, and creative, which repairs and generates more detail.

Runway is now served on OpenRouter with Aleph 2.0 and Gen-4.5 behind one API. Aleph 2.0 edits existing footage by changing a single frame and propagating that direction across the video, while Gen-4.5 generates new video with tighter prompt adherence and motion quality.

Voyage Code 4, MongoDB's code retrieval embedding model, is now available through OpenRouter with Matryoshka embeddings at 256, 512, 1024 and 2048 dimensions and multiple quantization options, giving coding agents a tunable index size.

Recraft V4 Styles is on OpenRouter in four variants: 1K raster, 2K Pro, and editable SVG vector versions of each. Every request takes 1 to 10 reference images and generates new work in their style, matching texture, composition, color and rendering without fine-tuning.

Ori Harness configures Codex, Claude Code, OpenCode, and Hermes to run on OpenRouter with settings tuned per harness. It spreads Opus traffic across eight endpoints on Anthropic, Bedrock, Vertex, and Azure, and flips a Claude Code setting that nearly halves the system prompt.

DeepSeek charges different rates by hour of day. OpenRouter now shows the peak and off-peak windows in your local timezone on the model page and factors the schedule into which endpoint it sources, with the V4 Flash 0731 endpoint currently showing a 64 percent discount.

OpenRouter benchmarked Exa, Parallel, Perplexity and native search across four models and four search depths. Budget mattered most: 1 to 25 turns roughly doubled BrowseComp scores. Model choice moved results ~15 points, engine choice ~10. Incorrect runs averaged 19.7 searches against 10.3 for correc

Grok 4.6 is routable on OpenRouter, with gains over Grok 4.5 claimed on GPDVal-AA V2 and pricing unchanged at $2 per million input tokens and $6 per million output.

A week of Ori Eval fixes: failures now report the cause and the serving provider, crashed runs count as failures rather than wrong answers, ori eval --pilot estimates full-run cost within 5%, the tool warns when judge and candidate share a family, and headless runs work without a browser login.

Sakana AI's Namazu is routable on OpenRouter. Built on Kimi K2.6, it pairs reasoning with integrated web search and code execution and is specialized for Japanese cultural and business context.

Ori Eval runs inside your coding agent, finds where the codebase calls models, asks what you optimize for, then writes an eval and ranks candidate models against it. Assertions cover which tools were called, which were avoided, and whether the run completed, with LLM-as-judge for open-ended answers.

MiniMax's open-weights Hailuo H3 is on OpenRouter: a lightweight video model taking text, image, audio and video input with native audio-visual output, built for instruction-led edits, text and brand rendering, and video-to-video motion transfer. Up to 12 reference files per request.

OpenAI cut prices on GPT-5.6 Terra and Luna, and OpenRouter's 50% exclusive discount stacks on top: Luna lands at $0.1/M input and $0.6/M output, Terra at $1/M input and $6/M output, pushing Luna further onto the cost-quality frontier.

Ox Alpha is an anonymous frontier model on OpenRouter aimed at efficient coding and sustained agentic work, with a 1M token context and text, image and video input. It is free during the stealth window, and the provider does not train on prompts or completions.

The upgraded Auto Router classifies each prompt into roughly 30 task types, then ranks models by the community's real spend share for that task over the prior 7 days, filtered by your cost tier. It keeps multi-turn conversations on one model to avoid cache rebuilds.