Kimi K2.7 Code
MoonshotReleased Jun 12, 2026Open weights · Modified MITkimi-k2.7-code
- Intelligence (Epoch)
- 150 · 40th of 111
- Reasoning
- yes
- Knowledge cutoff
- April 2026
- Context
- 262K tokens
- Max output
- 64k tokens
- Input
- text, image, video
- API features
- tools, structured outputs, JSON mode, seed, logprobs
- Moonshot's API
- yes, as kimi-k2.7-code
Read Oct 23, 06:00 UTC from the hosts' published offers and OpenRouter, spec sheet from Moonshot · cutoff: aiknowledgecutoff.com
The cheapest host, StreamLake, serves a weights precision we can't read (host terms). Cheapest at the maker's precision and context: Inceptron, $4.02 blended. Weights and context per host
Intelligence for the price
12 models score higher for the same price or less; it ranks 13 of the 72 models at its price or below.
Epoch Capabilities Index against the lowest input + output price per M, log scale · prices Oct 23. Score: Epoch AI, CC BY 4.0, 150 with an interval of 148 to 152.
Against its rivals
The best of its price range and the top score
- Kimi K2.7 Codethis model$3.71150
- Claude Opus 5.5top score$24.00169
- Gemini 3.7 Flashsame price range$4.50158
- Gemini 3.8 Flashsame price range$4.50157
Same level for 20× less: GLM-5.3-Flash at StreamLake, $0.19 per M input + output · their Epoch intervals overlap (153 against 150).
Bars: Epoch Capabilities Index (Epoch AI, CC BY 4.0), drawn from 100, below the lowest score in the catalogue · price: lowest input + output per M.
Speed by provider
Output speed of Kimi K2.7 Code at each host, fastest first
- Inceptron98 tok/s0.73 s to first token
- CoreWeave75 tok/s0.57 s to first token
- Nebius68 tok/s1.07 s to first token
- Cloudflare56 tok/s0.73 s to first token
- StreamLake54 tok/s0.80 s to first token
- Moonshot AI53 tok/s1.05 s to first token
- Alibaba47 tok/s0.86 s to first token
- Venice43 tok/s2.32 s to first token
- SiliconFlow29 tok/s0.91 s to first token
- GMICloud2 tok/s51.5 s to first token
Speed: OpenRouter, median of the last 30 min, read Oct 23, 06:00 UTC · a snapshot of real traffic before the daily read, not a daily average · bars from zero.
Run it yourself
open weights, Modified MIT · too large for the 6 workstations in our guideSelf-host guide ›- Parameters
- 1000B
- 32B active per token (MoE)
- Memory needed
- 600 GB
- in INT4 · 1200 GB in INT8
- Smallest cloud setup
- 8× H100
- 640 GB · $17.20/h, about $12,556 a month
- Worth it above
- 3B tokens
- a month, against the cheapest API at $3.71 per M
| Quantization | Memory | Smallest cloud GPU | Quality |
|---|---|---|---|
| INT4 (Q4) | 600 GB | 8× H100 · $17.20/h | smallest footprint, minor quality loss |
| INT8 (Q8) | 1200 GB | multi-node, over 640 GB | near-lossless |
| FP16 (full) | 2400 GB | multi-node, over 640 GB | reference quality |
At home: 600 GB in INT4 is beyond any desktop or Mac in our machines table, so it means a multi-GPU server.
Memory ≈ parameters × bytes per parameter × 1.2 · cloud price: on-demand rate, running 24/7 · builds: GGUF, AWQ, MLX, GPTQ, Ollama · break-even calculator
| Provider | Input | Output | Cache ↑ | tok/s | First token | Uptime 24h | Checked |
|---|---|---|---|---|---|---|---|
| DeepInfrafp4 · 256K ctx | $0.68 | $3.40 | $0.14 | — | — | — | Sep 29 |
| StreamLakeprecision not read: host terms · 250K ctx · 32K out−25% | $0.71 | $3.00 | $0.14 | 54 | 0.80 s | 100% | Oct 23 |
| CoreWeaveint4 · 256K ctx | $0.71 | $3.50 | $0.15 | 75 | 0.57 s | 100% | Oct 23 |
| Nebiusfp4 · 256K ctx | $0.95 | $4.00 | $0.15 | 68 | 1.07 s | 97.9% | Oct 23 |
| Veniceint4 · 250K ctx · 64K out | $0.75 | $3.50 | $0.16 | 43 | 2.32 s | 100% | Oct 23 |
| BaseTenprecision not stated · 256K ctx | $0.95 | $4.00 | $0.16 | — | — | 100% | Sep 27 |
| SiliconFlowfp8 · 256K ctx | $0.86 | $3.80 | $0.18 | 29 | 0.91 s | 100% | Oct 23 |
| Inceptronint4 · 256K ctx | $0.67 | $3.35 | $0.18 | 98 | 0.73 s | 99.9% | Oct 23 |
| Novita AIint4 · 256K ctx | $0.95 | $4.00 | $0.19 | — | — | — | Oct 23 |
| GMICloudfp8 · 256K ctx | $0.95 | $4.00 | $0.19 | 2 | 51.5 s | 100% | Oct 23 |
| Alibabafp8 · 256K ctx · 16K out | $0.95 | $4.00 | $0.19 | 47 | 0.86 s | 93.1% | Oct 23 |
| Together AIprecision not stated · 256K ctx | $0.95 | $4.00 | $0.19 | — | — | — | Sep 8 |
| Cloudflareprecision not stated · 256K ctx | $0.95 | $4.00 | $0.19 | 56 | 0.73 s | 100% | Oct 23 |
| Moonshot AIint4 · 256K ctx | $0.95 | $4.00 | $0.19 | 53 | 1.05 s | 100% | Oct 23 |
| Phalaprecision not read: host terms · 256K ctx | $0.95 | $4.00 | $0.19 | — | — | — | Oct 22 |
10 of 15 providers · prices per million tokens, read Oct 23, 06:00 UTC · speed: OpenRouter, median of the last 30 min, read Oct 23, 06:00 UTC · uptime: OpenRouter, last 24 h · −N%: a discount off the host's list price · under each host: weights precision, context and max output as served, the maker's value in brackets when the host's differs
| Provider | Weights | Context | Max output | Tools | Structured outputs | Reasoning |
|---|---|---|---|---|---|---|
| StreamLake | not read: host terms | 250K | 32K | yes | yes | yes |
| Inceptron | int4 | 256K | not stated by the host | yes | yes | yes |
| DeepInfra | fp4 | 256K | not stated by the host | — | — | — |
| CoreWeave | int4 | 256K | not checked yet | yes | yes | yes |
| Venice | int4 | 250K | 64K | yes | yes | yes |
| SiliconFlow | fp8 | 256K | not read: host terms | yes | no | yes |
| Novita AI | int4 | 256K | not stated by the host | yes | yes | yes |
| BaseTen | not stated by the host | 256K | not stated by the host | — | — | — |
| GMICloud | fp8 | 256K | not read: host terms | yes | yes | yes |
| Alibaba | fp8 | 256K | 16K | yes | yes | yes |
| Together AI | not stated by the host | 256K | not stated by the host | — | — | — |
| Cloudflare | not stated by the host | 256K | not stated by the host | yes | yes | yes |
| Moonshot AI | int4 | 256K | not checked yet | yes | yes | yes |
| Nebius | fp4 | 256K | not stated by the host | yes | yes | yes |
| Phala | not read: host terms | 256K | not read: host terms | — | — | — |
Weights, context and max output: read at each host's own catalogue or docs where its terms allow, otherwise as the host declares to OpenRouter; hover a cell for its source and date · tools, structured outputs, reasoning: as declared to OpenRouter · not verified by us
Kimi K2.7 Code spec sheet, field by field, with sources
Kimi K2.7 Code spec sheetmaker-declared
What Moonshot declares about the model itself — read from its own weights repository and documentation, not from any host. What each endpoint actually serves is in the pricing above: the two can differ, and both are true.
- Context window
- 262,144 tokens
- Parameters
- 1,026,879,376,368≈ 1.03T
- Active experts
- 8 of 384 active per token
- Weight precision
- not published by the maker
- License
- Modified MIT — open weights, conditionalcuratedScope: the licence on the published weights, and nothing else. It is not the maker's acceptable-use policy, which is a separate document, and it is not the contract you sign with whichever host you call — hosts set their own terms, which this index does not read.
- Knowledge cutoff
- not published by the maker≈ April 2026 third-party estimate — not the maker's figurecurated
- Modalities
- text · image · video
- Max output (official)
- to verify
- Release date
- to verify≈ June 12, 2026 third-party catalogue — not the maker's figure
- Maker lifecycle status
- to verify
Also declared in the weights
- Architecture class
- KimiK25ForConditionalGeneration
- Layers
- 61
- Vocabulary
- 163,840 tokens
- Weights first published
- June 11, 2026
- Weights last updated
- June 15, 2026
moonshotai/Kimi-K2.7-Code · config.json @ 74797c9 · read 2026-10-08 · 5 of 10 fields published by the maker
| Date | Provider | Input | Output | Change |
|---|---|---|---|---|
| Sep 30 | Inceptron | $0.66 → $0.67 | $3.30 → $3.35 | +2% |
| Sep 25 | Inceptron | $0.71 → $0.66 | $3.30 → $3.30 | −1% |
| Sep 23 | Inceptron | $0.71 → $0.71 | $3.21 → $3.30 | +2% |
| Sep 7 | Novita AI | $0.91 → $0.95 | $3.84 → $4.00 | +4% |
| Sep 5 | Moonshot AI | $1.90 → $0.95 | $8.00 → $4.00 | −50% |
Every published change of input or output price, newest first · change on input + output · full series: pricing-history.csv.
| Provider | Chatbot 1,000 in / 500 out | RAG / long context 8,000 in / 500 out | Batch summarize 4,000 in / 1,000 out |
|---|---|---|---|
| StreamLake | $2.21 | $7.20 | $5.85 |
| Inceptron | $2.35 | $7.04 | $6.03 |
| DeepInfra | $2.38 | $7.14 | $6.12 |
| CoreWeave | $2.46 | $7.43 | $6.34 |
| Venice | $2.50 | $7.75 | $6.50 |
| SiliconFlow | $2.76 | $8.77 | $7.24 |
| Novita AI | $2.95 | $9.60 | $7.80 |
| BaseTen | $2.95 | $9.60 | $7.80 |
| GMICloud | $2.95 | $9.60 | $7.80 |
| Alibaba | $2.95 | $9.60 | $7.80 |
| Together AI | $2.95 | $9.60 | $7.80 |
| Cloudflare | $2.95 | $9.60 | $7.80 |
| Moonshot AI | $2.95 | $9.60 | $7.80 |
| Nebius | $2.95 | $9.60 | $7.80 |
| Phala | $2.95 | $9.60 | $7.80 |
What each host bills for 1,000 requests of a representative job
Your own workload
Other Moonshot models · all its models, hosts and history ›
- Kimi K2.618 providers · from $2.40
- Kimi K317 providers · from $11.35
- Kimi K2.58 providers · from $2.70
- Kimi K2 Thinking2 providers · from $3.10
Head-to-head: Qwen3 Coder Next (Alibaba) vs Kimi K2.7 Code (Moonshot AI)