DeepSeek-R1 vs Kimi K2.7 Code vs Qwen3 235B-A22B
Kimi K2.7 Code comes out ahead, 64 to 51 and 51 on our weighted score, though DeepSeek-R1 is 31% cheaper per token.
DeepSeekDeepSeek-R1
51/100- ECI139.0
- Price$0.70 / $2.60
- Context128K
- Our pick
Moonshot AIKimi K2.7 Code
64/100- ECI150.0
- Price$0.95 / $4.00
- Context262K
Alibaba (Qwen)Qwen3 235B-A22B
51/100- ECI139.4
- Price$0.70 / $2.80
- Context131K
Kimi K2.7 Code is our pick
Kimi K2.7 Code is the better all-round choice, scoring 64/100 against Qwen3 235B-A22B (51) and DeepSeek-R1 (51). It leads on capability, inputs & features and context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityKimi K2.7 CodeCapabilities Index (ECI): Kimi K2.7 Code 150.0 · Qwen3 235B-A22B 139.4 · DeepSeek-R1 139.0
- Lowest priceDeepSeek-R1DeepSeek-R1 $1.18 · Qwen3 235B-A22B $1.23 · Kimi K2.7 Code $1.71 per 1M tokens (3:1 blend)
- Longest contextKimi K2.7 CodeKimi K2.7 Code 262,144 · Qwen3 235B-A22B 131,072 · DeepSeek-R1 128,000 tokens
- Widest inputsKimi K2.7 CodeDeepSeek-R1: Text · Kimi K2.7 Code: Text, Images, Video · Qwen3 235B-A22B: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | DeepSeek-R1 | Kimi K2.7 Code | Qwen3 235B-A22B |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 64 | 78 | 65 |
| Price | 25% | 47 | 39 | 46 |
| Inputs & features | 15% | 35 | 80 | 35 |
| Context window | 10% | 24 | 37 | 24 |
| Overall | 100% | 51/100 | 64/100 | 51/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | DeepSeek-R1DeepSeek | Kimi K2.7 CodeMoonshot AI | Qwen3 235B-A22BAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 139.0 | 150.0 (best) | 139.4 |
| ECI rank | #104 of 148 | #49 of 148 (best) | #103 of 148 |
| GPQA DiamondGraduate-level science questions | 71.7% | 87.9% (best) | 70.7% |
| FrontierMath Tiers 1–3Research-level mathematics | — | 54.0% | — |
| OTIS Mock AIME 2024–2025Competition mathematics | 53.3% | 95.6% (best) | — |
| SimpleQA VerifiedShort factual questions | — | 36.5% | — |
| Price per million tokens | |||
| Input | $0.70 (best) | $0.95 | $0.70 (best) |
| Output | $2.60 (best) | $4.00 | $2.80 |
| Cached input | — | $0.19 | — |
| Blended (3:1) | $1.18 (best) | $1.71 | $1.23 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 11 providers | Official Moonshot AI API | Official Alibaba API |
| Limits | |||
| Context window | 128,000 tokens | 262,144 tokens (best) | 131,072 tokens |
| Max output | 32,768 tokens | 262,144 tokens (best) | 16,384 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | Yes | No |
| Reasoning | Yes | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | No |
| Availability | |||
| Weights | Open | Open | Open |
| API model ID | — | kimi-k2.7-code | qwen3-235b-a22b |
| API providers | 12 | 51 (best) | 7 |
| Released | Jan 20, 2025 | Jun 12, 2026 | Apr 28, 2025 |
| Knowledge cutoff | Jul 2024 | Jan 2025 | Apr 2025 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- DeepSeek-R1$12.20
- Kimi K2.7 Code$17.50
- Qwen3 235B-A22B$12.60
Which should you choose?
Which is better: DeepSeek-R1, Kimi K2.7 Code or Qwen3 235B-A22B?
Kimi K2.7 Code is the better all-round choice, scoring 64/100 against Qwen3 235B-A22B (51) and DeepSeek-R1 (51). It leads on capability, inputs & features and context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, DeepSeek-R1, Kimi K2.7 Code or Qwen3 235B-A22B?
DeepSeek-R1 is cheaper at $0.70 input / $2.60 output per million tokens (median across 11 API providers). Qwen3 235B-A22B costs $0.70 input / $2.80 output per million tokens (official Alibaba API price); Kimi K2.7 Code costs $0.95 input / $4.00 output per million tokens (official Moonshot AI API price). At a typical mix of three input tokens to one output token, that is $1.18 per million tokens for DeepSeek-R1 versus $1.23 for Qwen3 235B-A22B (1× as much) and $1.71 for Kimi K2.7 Code (1.5× as much).
Which scores higher on benchmarks?
Kimi K2.7 Code scores higher on the Capabilities Index (ECI): Kimi K2.7 Code 150.0 (#49 of 148), Qwen3 235B-A22B 139.4 (#103 of 148) and DeepSeek-R1 139.0 (#104 of 148). Their confidence ranges do not overlap (148.1–151.8 vs 135.2–140.8), so the gap is a real one. On individual benchmarks: GPQA Diamond — Kimi K2.7 Code 87.9%, DeepSeek-R1 71.7%, Qwen3 235B-A22B 70.7%.
Which is better for coding?
There are no published SWE-bench Verified results for DeepSeek-R1, Kimi K2.7 Code and Qwen3 235B-A22B yet, so there is no like-for-like coding score. On overall capability, Kimi K2.7 Code leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.
Which has the bigger context window?
Kimi K2.7 Code has the largest context window at 262,144 tokens, against 131,072 for Qwen3 235B-A22B and 128,000 for DeepSeek-R1. Maximum output per response: DeepSeek-R1 up to 32,768, Kimi K2.7 Code up to 262,144, Qwen3 235B-A22B up to 16,384 tokens.
Which can read images, PDFs, audio or video?
DeepSeek-R1 accepts text; Kimi K2.7 Code accepts text, images and video; Qwen3 235B-A22B accepts text. Kimi K2.7 Code handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights, so you can self-host them.
Which is newer?
Kimi K2.7 Code is the newest, released Jun 12, 2026. Qwen3 235B-A22B came out Apr 28, 2025; DeepSeek-R1 came out Jan 20, 2025. Knowledge cutoff: DeepSeek-R1 Jul 2024, Kimi K2.7 Code Jan 2025, Qwen3 235B-A22B Apr 2025.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Oct 4, 2026.
