Kimi K2 Thinking vs GPT-5-Codex vs Qwen3.5 Plus
Qwen3.5 Plus comes out ahead, 59 to 42 and 42 on our weighted score, and it is the cheaper option too.
Moonshot AIKimi K2 Thinking
42/100- ECI146.0
- Price$0.60 / $2.50
- Context262K
OpenAIGPT-5-Codex
42/100- ECI—
- Price$1.25 / $10.00
- Context400K
- Our pick
Alibaba (Qwen)Qwen3.5 Plus
59/100- ECI146.8
- Price$0.40 / $2.40
- Context1M
Qwen3.5 Plus is our pick
Qwen3.5 Plus is the better all-round choice, scoring 59/100 against GPT-5-Codex (42) and Kimi K2 Thinking (42). It leads on price and context window. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceQwen3.5 PlusQwen3.5 Plus $0.90 · Kimi K2 Thinking $1.07 · GPT-5-Codex $3.44 per 1M tokens (3:1 blend)
- Longest contextQwen3.5 PlusQwen3.5 Plus 1,000,000 · GPT-5-Codex 400,000 · Kimi K2 Thinking 262,144 tokens
- Widest inputsQwen3.5 PlusKimi K2 Thinking: Text · GPT-5-Codex: Text, Images · Qwen3.5 Plus: Text, Images, Video
- Self-hostingKimi K2 ThinkingPublishes downloadable weights
| Measure | Weight | Kimi K2 Thinking | GPT-5-Codex | Qwen3.5 Plus |
|---|---|---|---|---|
| Price | 50% | 48 | 24 | 52 |
| Inputs & features | 30% | 35 | 70 | 70 |
| Context window | 20% | 37 | 44 | 60 |
| Overall | 100% | 42/100 | 42/100 | 59/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Kimi K2 ThinkingMoonshot AI | GPT-5-CodexOpenAI | Qwen3.5 PlusAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 146.0 | — | 146.8 (best) |
| ECI rank | #72 of 148 | — | #65 of 148 (best) |
| GPQA DiamondGraduate-level science questions | 84.2% | — | 84.9% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | 83.1% | — | 86.7% (best) |
| SimpleQA VerifiedShort factual questions | — | — | 25.4% |
| Price per million tokens | |||
| Input | $0.60 | $1.25 | $0.40 (best) |
| Output | $2.50 | $10.00 | $2.40 (best) |
| Cached input | — | — | — |
| Blended (3:1) | $1.07 | $3.44 | $0.90 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 10 providers | Median of 3 providers | Official Alibaba API |
| Limits | |||
| Context window | 262,144 tokens | 400,000 tokens | 1,000,000 tokens (best) |
| Max output | 262,144 tokens (best) | 128,000 tokens | 65,536 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | Yes | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | No |
| Availability | |||
| Weights | Open | Proprietary | Proprietary |
| API model ID | — | — | qwen3.5-plus |
| API providers | 10 (best) | 3 | 10 (best) |
| Released | Nov 6, 2025 | Sep 15, 2025 | Feb 16, 2026 |
| Knowledge cutoff | Aug 2024 | Sep 30, 2024 | 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.
- Kimi K2 Thinking$11.00
- GPT-5-Codex$32.50
- Qwen3.5 Plus$8.80
Which should you choose?
Which is better: Kimi K2 Thinking, GPT-5-Codex or Qwen3.5 Plus?
Qwen3.5 Plus is the better all-round choice, scoring 59/100 against GPT-5-Codex (42) and Kimi K2 Thinking (42). It leads on price and context window. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, Kimi K2 Thinking, GPT-5-Codex or Qwen3.5 Plus?
Qwen3.5 Plus is cheaper at $0.40 input / $2.40 output per million tokens (official Alibaba API price). Kimi K2 Thinking costs $0.60 input / $2.50 output per million tokens (median across 10 API providers); GPT-5-Codex costs $1.25 input / $10.00 output per million tokens (median across 3 API providers). At a typical mix of three input tokens to one output token, that is $0.90 per million tokens for Qwen3.5 Plus versus $1.07 for Kimi K2 Thinking (1.2× as much) and $3.44 for GPT-5-Codex (3.8× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Kimi K2 Thinking has an ECI of 146.0, GPT-5-Codex has not been scored yet and Qwen3.5 Plus has an ECI of 146.8.
Which is better for coding?
There are no published SWE-bench Verified results for Kimi K2 Thinking, GPT-5-Codex and Qwen3.5 Plus yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
Qwen3.5 Plus has the largest context window at 1,000,000 tokens, against 400,000 for GPT-5-Codex and 262,144 for Kimi K2 Thinking. Maximum output per response: Kimi K2 Thinking up to 262,144, GPT-5-Codex up to 128,000, Qwen3.5 Plus up to 65,536 tokens.
Which can read images, PDFs, audio or video?
Kimi K2 Thinking accepts text; GPT-5-Codex accepts text and images; Qwen3.5 Plus accepts text, images and video. Qwen3.5 Plus handles the widest range of inputs.
Are any of these open source?
Kimi K2 Thinking publishes its weights and can be self-hosted; GPT-5-Codex and Qwen3.5 Plus is proprietary.
Which is newer?
Qwen3.5 Plus is the newest, released Feb 16, 2026. Kimi K2 Thinking came out Nov 6, 2025; GPT-5-Codex came out Sep 15, 2025. Knowledge cutoff: Kimi K2 Thinking Aug 2024, GPT-5-Codex Sep 30, 2024, Qwen3.5 Plus 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.
