Kimi K3 moonshotai/kimi-k3
Served by 18 providers. Cheapest: Inference.net at $4.31 blended /1M tokens. The most expensive listing costs 1.7× as much for the same model.
| Provider | Type | Input /1M | Output /1M | Cached input | Blended | Context | Notes | Source |
|---|---|---|---|---|---|---|---|---|
| Inference.net moonshotai/kimi-k3 |
Inference hosts | $2.10 | $10.95 | $0.23 | $4.31 | 1.05M | https://inference.net/models shows the same figures; separate "kimi-k3-fast" variant at $4.50/$22.50 |
✓ src 2026-09-18 |
| OpenRouter moonshotai/kimi-k3 |
LLM routers | $2.10 | $10.95 | $0.23 | $4.31 | 1.05M | api modality text+image+video->text |
✓ src 2026-09-18 |
| DigitalOcean kimi-k3 |
Inference hosts | $2.55 | $12.95 | $0.285 | $5.15 | 1.05M |
✓ src src2 2026-09-18 |
|
| SiliconFlow | Inference hosts | $2.70 | $13.50 | $0.27 | $5.40 | — | context shown on pricing page as "1049K" |
✓ src 2026-09-18 |
| DeepInfra moonshotai/Kimi-K3 |
Inference hosts | $2.85 | $14.25 | $0.285 | $5.70 | 1.05M |
✓ src src2 2026-09-18 |
|
| Atlas Cloud moonshotai/kimi-k3 |
Inference hosts | $3 | $15 | $0.30 | $6 | 1.05M | int4 |
✓ src 2026-09-18 |
| Baseten | Inference hosts | $3 | $15 | $0.30 | $6 | — | Model API id not shown on pricing page. |
◐ src 2026-09-18 |
| Chutes moonshotai/Kimi-K3-TEE |
Inference hosts | $3 | $15 | $0.30 | $6 | 1.05M | mxfp4 TEE (confidential compute); upstream weights moonshotai/Kimi-K3 |
✓ src src2 2026-09-18 |
| Fireworks AI accounts/fireworks/models/kimi-k3 |
Inference hosts | $3 | $15 | $0.30 | $6 | — | batch −50% Priority $3.75/$0.375/$18.75; Fast $4.50/$0.45/$22.50; US-only $4.50/$0.45/$22.50 |
✓ src 2026-09-18 |
| GMI Cloud Inference Engine moonshotai/kimi-k3 |
Inference hosts | $3 | $15 | — | $6 | — | from GMI console public pricing API (console is client-rendered; docs say prices live in console Model Hub) |
✓ src src2 2026-09-18 |
| Microsoft Foundry (Azure AI Foundry / Azure OpenAI) Kimi-K3 |
Hyperscaler AI platforms | $3 | $15 | $0.30 | $6 | — | Global Standard, Azure Retail Prices API meter 'FW Kimi-K3 Gl Inp Tokens' (eastus); API meter is per 1K tokens, multiplied by 1000 (unit conversion only); sold under Azure Fireworks Models product (Fireworks-served, Azure-billed) |
✓ src src2 2026-09-18 |
| Moonshot AI (Kimi API Platform) kimi-k3 |
Model labs | $3 | $15 | $0.30 | $6 | 1.05M | No Batch API. 2.8T-parameter model with native visual understanding per model list. |
✓ src 2026-09-18 |
| Novita AI moonshotai/kimi-k3 |
Inference hosts | $3 | $15 | $0.30 | $6 | 1.05M |
✓ src src2 2026-09-18 |
|
| Parasail | Inference hosts | $3 | $15 | $0.30 | $6 | — |
✓ src 2026-09-18 |
|
| Requesty moonshot/kimi-k3 |
LLM routers | $3 | $15 | $0.30 | $6 | 1.05M | mxfp4 upstream route moonshot; cache write $15/M; geo global; 5 other upstream routes listed (input $3-$3/M); API list price; Requesty's 5% markup on model cost (pricing page) is not included |
✓ src 2026-09-18 |
| Together AI moonshotai/Kimi-K3 |
Inference hosts | $3 | $15 | $0.30 | $6 | 1.05M |
✓ src 2026-09-18 |
|
| Amazon Bedrock moonshotai.kimi-k3 |
Hyperscaler AI platforms | $3.30 | $16.50 | $0.33 | $6.60 | 1M | batch −50% In-region us-east-1. Global cross-region: $3.00/$15.00/$0.30 cache read |
✓ src src2 2026-09-18 |
| Venice.ai API kimi-k3 |
Inference hosts | $3.75 | $18.75 | $0.375 | $7.50 | 1M |
✓ src src2 2026-09-18 |
✓ from the provider's own pricing page · ◐ partly confirmed there · 2° third-party source. Router prices may exclude the router's own fees; see each router's page.