MoonshotaiReasoningReleased 16 July 2026
Kimi K3 pricing, context and benchmark
Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI.
1.05M
Input context
944K
Output context
$0.182
Official input
$11.005
Official output
n/a
AA index
01 · Model Overview
Everything that matters, in one table
Context, pricing, cache, modalities and tooling, from the model registry and Artificial Analysis.
| Model | Kimi K3 |
| Model ID | moonshotai/kimi-k3 |
| Provider | Moonshotai |
| Release date | 16 July 2026 |
| Status | Generally available |
| Input context | 1,048,576 tokens |
| Output context | 943,718 tokens |
| Official input pricing | $0.182 per 1M tokens |
| Official output pricing | $11.005 per 1M tokens |
| Cache read pricing | $0.3 per 1M tokens (90% off input) |
| Market input pricing | $0.153 per 1M (cheapest route) |
| Market output pricing | $16 per 1M (cheapest route) |
| Input modalities | text · image · video |
| Output modalities | text |
| Reasoning control | Not exposed |
| Tool support | Text generation only |
| Tokenizer | Other |
| Provider routes | 8 endpoints |
| Artificial Analysis index | Not published |
| Speed | 42 tokens/s, 67.87s to first token |
| Knowledge cutoff | Not publicly stated |
Input and output support
| Modality | Input | Output | Billed at |
|---|---|---|---|
| Text | ✓ yes | ✓ yes | $0.182 in |
| Image | ✓ yes | no | $0.182 in |
| Audio | no | no | not supported |
| Video | ✓ yes | no | $0.182 in |
| File (PDF) | no | no | not supported |
02 · Overview
What is 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...
Kimi K3 carries a 1.05M token window, uses the Other tokenizer and is served by Moonshotai. On Hoodgen it sits in the Reasoning tier of the model menu next to its siblings, so you can switch mid conversation without losing the thread.
03 · Release
Release and what changed from the previous version
Released16 July 2026
| Property | Kimi K2.7 Code | Kimi K3 | Change |
|---|---|---|---|
| Released | 2026 | 2026 | Same release window |
| Context | 262K | 1.05M | Changed |
| Max output | 236K | 944K | Changed |
| Input price | $0.6 | $0.182 | Cheaper |
| Output price | $3 | $11.005 | Higher |
| Input modalities | text, image | text, image, video | Unchanged |
| AA index | n/a | n/a | Newer benchmark run |
04 · Capabilities
Features and capabilities
◉
Vision
Screenshots, charts and photographed documents are read at the text token rate.
♪
Audio and video
Calls, meetings and screen recordings can be summarised in one pass.
⚡
Prompt caching
Cached input costs $0.3 per million instead of $0.182, so stable prefixes get cheaper.
07 · Use cases
What can you do with Kimi K3?
Debugging and root cause
Multi file failures where the model has to hold a causal chain before answering.
Architecture and planning
Design docs, migration plans and trade off analysis over long context.
Data analysis
Spreadsheets and logs read in one pass, then summarised with the numbers quoted back.
Agentic coding
Repo wide changes that need plan, edit and verify loops.
08 · Cost levers
Prompt caching and how to cut the bill
Cached input bills at $0.3 per million instead of $0.182, a 90% discount. Keep the system prompt and documents first and the changing question last, so every call hits the cache.
09 · Pricing
Kimi K3 pricing
Official pricing
List price
$0.182 / 1M in
$11.005 / 1M out
- $0.3 cached input
- 1.05M context
- text, image, video input
Cheapest route
$0.153 / 1M in
$16 / 1M out
- 8 routes on OpenRouter
- What Hoodgen routes to first
- Same model, lower bill
Market pricing, every route
| Provider | Tier | Input | Output | Cache read | Uptime 24h |
|---|---|---|---|---|---|
| Wafer | priority | $0.153 | $16 | $0.152 | 100.0% |
| Relace | standard | $0.182 | $9.339 | $0.18 | 100.0% |
| Wafer | standard | $0.192 | $16 | $0.182 | 100.0% |
| InferenceNet | standard | $0.234 | $15 | $0.195 | 100.0% |
| InferenceNet | standard | $0.234 | $15 | $0.195 | 100.0% |
| Wafer | standard | $0.276 | $9.6 | $0.13 | 100.0% |
| Sail Research | standard | $0.276 | $13.264 | $0.138 | 100.0% |
| Sail Research | standard | $0.276 | $13.264 | $0.138 | 100.0% |
API pricing and cost calculator
| Rate | Per call | Total | Credits |
|---|---|---|---|
| Cheapest route | $0.0270 | $13.52 | 12848 |
| List price | $0.0216 | $10.78 | 10245 |
10 · Developers
API and developer information
| Model ID | moonshotai/kimi-k3 |
| Endpoint | https://openrouter.ai/api/v1/chat/completions |
| Parameters | prompt and max_tokens |
| Streaming | Yes, server sent events |
| Modality string | text+image+video->text |
| Max prompt tokens | 1,048,576 |
| Max completion tokens | 943,718 |
curl https://openrouter.ai/api/v1/chat/completions \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "moonshotai/kimi-k3",
"messages": [{ "role": "user", "content": "Summarise this document." }],
"max_tokens": 1024
}'11 · Limitations
Model limitations
| Text only output | It cannot generate images, audio or video. Pair it with an image model on Hoodgen when you need visuals. |
| Output cap | 943,718 tokens per request. Long deliverables have to be chunked. |
| Benchmarks not published | Artificial Analysis does not publish an Intelligence Index for this release, so no benchmark score is claimed here. |
| Knowledge cutoff | The provider does not state a fixed cutoff for every release. Verify very recent events against search. |
12 · Comparison
Kimi K3 vs other Moonshotai models
| Model | Context | Input /1M | Output /1M | AA index | Released |
|---|---|---|---|---|---|
| Kimi K3 | 1.05M | $0.182 | $11.005 | n/a | 2026 |
| Kimi K2 0905 (exacto) | n/a | $0.6 | $2.5 | n/a | 2026 |
| Kimi K2.7 Code | 262K | $0.6 | $3 | n/a | 2026 |
13 · Privacy
Privacy and data policy
- On Hoodgen, prompts and attachments answer that request and stay in your conversation history so you can reopen it. They are not used to train models.
- Requests route through OpenRouter to Moonshotai and are covered by that provider's API terms.
- Do not paste secrets or personal data you would not send to any hosted model.
14 · FAQ
Kimi K3 questions, answered
Kimi K3 is a reasoning model from Moonshotai, released 16 July 2026. Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI.
✦
Start chatting with Kimi K3
Free to try on Hoodgen AI, with a live credit estimate before every send. No API key, no setup.