The Orchestrator › Models › Moonshot AI
Moonshot AI: MoonshotAI: Kimi K3
Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...
Specifications
| Provider | Moonshot AI |
| Model ID | moonshotai/kimi-k3 |
| Context window | 1.0M tokens (1,048,576) |
| Modalities | text + image + video → text |
| Tokenizer | Other |
| Router sources | OpenRouter, TrustedRouter, Vercel AI Gateway, Cloudflare |
| Released | 2026-07-16 |
| Tracked since | 2026-11-14 |
| Popularity | #17 by usage · peak 8.9% spend share in an OpenRouter category |
| Reasoning | On by default |
Pricing (per million tokens)
| Variant | Source | API ID | Input | Output | Context |
|---|---|---|---|---|---|
| standard | OpenRouter | moonshotai/kimi-k3 | $0.192 | $16.00 | 1.0M |
| batch | OpenRouter | moonshotai/kimi-k3:batch | $2.28 | $11.40 | 1.0M |
| standard | TrustedRouter | moonshotai/kimi-k3 | $0.38 | $10.02 | 1.0M |
| standard | Vercel AI Gateway | moonshotai/kimi-k3 | $3.00 | $15.00 | 1M |
| standard | Cloudflare | moonshotai/kimi-k3 | $3.00 | $15.00 | — |
Recorded price history
| Date | Input / 1M | Output / 1M | Context |
|---|---|---|---|
| 2026-11-14 | $3.00 | $15.00 | — |
| 2026-11-14 | $3.00 | $15.00 | 1M |
| 2026-11-14 | $0.38 | $10.02 | 1.0M |
| 2026-11-14 | $0.192 | $16.00 | 1.0M |
Capabilities
Vision (image input)Video inputAgentic tool useStructured outputsReasoning mode1M+ context windowControllable reasoningNo content storage (attested)Zero data retention
Trust flags come from TrustedRouter's per-provider attestations.
Supported parameters
frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p