93 models from 11 providers.

Anthropic’s fastest near-frontier model. 200k context, 64k output.

Previous flagship — 1M context. Still served alongside 4.7.

Top reasoning + agentic coding. 1M context, 128k output.

Balanced flagship — best speed/intelligence ratio. 1M context.

Cheaper prior-gen TTS — same controllability, narrower language set.

Mid-tier multimodal. Image + video + text input. 1M context.

Cheapest 1M-context tier. Great for high-volume agents.

Realtime conversational audio — bidirectional, sub-second latency. ~$0.005/min in, $0.018/min out.

Controllable TTS across 70+ languages, 200+ inline emotion tags. Audio output tokens.

Google’s flagship reasoning model + Computer Use. 1M context.

Original Nano Banana (Gemini 2.5 Flash Image). Fast, fun edits.

Nano Banana 2 (Gemini 3.1 Flash Image) — Pro features at Flash speed. $0.067/1K, $0.10/2K, $0.15/4K.

Nano Banana Pro (Gemini 3 Pro Image) — best fidelity, complex prompts, accurate text. $0.134/1K-2K, $0.24/4K.

Cinematic text/image-to-video, 720p–1080p, optional audio.

Veo 3.1 Fast — $0.10/s @720p, $0.12/s @1080p, $0.30/s @4K.

TTS with Instant Custom Voice (10s reference audio).

Gemini 3.1 Pro via GCP Vertex — SLA, region pinning, audit logs.

Highest-quality Imagen text-to-image with inpainting + editing.

Music generation up to 184s. Public preview.

Veo 3.1 via GCP Vertex.

Veo 3.1 Fast on Vertex — enterprise routing, $0.10/s @720p.

Text-or-audio in, text-or-audio out via /v1/chat/completions.

Token-priced TTS — ~$0.015/min generated audio.

ASR with optional speaker diarization. Same price as Whisper.

March 2026 release — sits between mini and 5.5. Strong cost/intel ratio.

GPT-5.4-class capability, fast + efficient. 400k context.

OpenAI’s smallest, cheapest model. High-volume simple tasks.

Flagship multimodal. 1M context. Computer use, MCP, hosted shell.

Highest-tier flagship — research-grade reasoning. No cached-input discount.

OpenAI image gen — flexible sizes, high-fidelity image input.

Bidirectional voice + text. ~$0.06/min audio in, $0.24/min audio out. Text in/out at $5/$20.

Cheap reasoning model — math, code, structured analysis at low cost.

Standard TTS — same voices, lower fidelity, half the cost of HD.

High-fidelity TTS — 6 preset voices. /v1/audio/speech endpoint.

Speech-to-text (ASR). Will be superseded by gpt-4o-transcribe.

Budget-friendly Grok — low TTFT, very cheap for high-volume tasks.

Long-context variant — 2M context window for very large corpora.

xAI’s May 2026 flagship — 1M context, multimodal.

Speech-to-text — multilingual, low-latency.

Flagship Qwen3 — Alibaba’s largest dense + MoE model.

Omni-modal: text + image + audio in/out. Cheapest realtime-class tier.

Vision-language flagship — strong OCR, document QA, GUI grounding.

Feb 2026 release — 397B params, 201 languages, 19× faster than prior gen.

Tiniest Qwen — extreme high-volume / on-device fallback.

Latest Hailuo text-to-video. Billed in units; ~$0.04/s @768p.

Hailuo 2.3 Fast — 30% cheaper than 2.3, same family.

Text-to-image, photoreal + stylized.

Older M-series — kept for reproducibility.

Prior-gen M-series — same base price, slightly weaker quality.

MiniMax’s newest self-iterating flagship for code + agents.

M2.7 with priority routing — 2× cost for lower TTFT.

Music generation from prompt — vocals + instrumentation.

Prior-gen HD voice — kept for reproducibility of pipelines.

HD TTS — 40 languages, accurate voice replication.

Latest TTS — Fluent LoRA voice cloning, prosodic naturalness across 40+ languages.

July 2025 K2 preview snapshot — scheduled for discontinuation 2026-05-25. (Not currently listed on OpenRouter — keeping prior estimate.)

September 2025 K2 preview snapshot — scheduled for discontinuation 2026-05-25.

Explicit thinking-mode K2 — chain-of-thought reasoning for math + code.

Moonshot’s agentic K2 with priority routing — fast TTFT, 128k context.

January 2026 multimodal release — cheaper than K2.6 with similar capability.

Moonshot’s April 2026 flagship — multimodal, 256k context.

Auto-routes to Moonshot’s current default (K2.6 as of 2026-05). Multimodal.

Legacy text-only Moonshot v1 long-context. (Not on OpenRouter — keeping prior flat pricing.)

Legacy alias → V4 Flash non-thinking mode. Sunsets 2026-07-24; migrate to deepseek-v4-flash.

Legacy alias → V4 Flash thinking mode. Sunsets 2026-07-24; migrate to deepseek-v4-flash with reasoning. (Not separately listed on OpenRouter — mirrors v4-flash.)

V4 default — 1M context, 384k max output.

V4 flagship reasoning — 1M context. OR pricing reflects 75%-off promo through 2026-05-31.

Zhipu’s text/image-to-video — 6s clips at 720p–1080p.

Free video generation tier.

Free image generation tier — generous rate limits.

Latest text-to-image — strong on Chinese text rendering + complex prompts.

Low-latency AirX variant — fastest GLM tier, smaller context.

Free tier — high-volume, simple tasks. Generous rate limits.

GLM 4 family flagship — 200k context, agentic + coding focused.

GLM 4.6 vision variant — multimodal input on the 4-family flagship.

Previous-gen flagship before GLM 5 — solid coding + reasoning at lower cost.

Lowest tier — high-volume simple tasks. Generous rate limits.

GLM 5 base flagship — released 2026-02-12, available on Pro and Max coding tiers.

March 2026 turbo variant — priority routing on the GLM 5 family.

Zhipu’s April 2026 flagship — SOTA on SWE-Bench Pro, 200k context, 128k max output.

GLM 5 family vision-capable turbo — multimodal with priority routing.

End-to-end voice + video understanding with singing + 2-min memory. Function calls supported.

Controllable + emotion-expressive zero-shot voice cloning. Open-sourced Dec 2025.

Claude Haiku 3.5 through AWS Bedrock US cross-region profile. Includes the documented 10% premium.

Claude Haiku 4.5 through AWS Bedrock Global application inference profile.

Claude Opus 4.1 through AWS Bedrock US cross-region profile. Includes the documented 10% premium.

Claude Opus 4.5 through AWS Bedrock Global application inference profile.

Claude Opus 4.6 through AWS Bedrock Global application inference profile.

Claude Opus 4.7 through AWS Bedrock Global application inference profile.

Claude Opus 4 through AWS Bedrock US cross-region profile. Includes the documented 10% premium.

Claude Sonnet 4 through AWS Bedrock Global application inference profile.

Claude Sonnet 4.5 through AWS Bedrock Global application inference profile.

Claude Sonnet 4.6 through AWS Bedrock Global application inference profile.