GLM-4.5-Air vs Qwen3 235B-A22B Thinking 2507 vs Codestral 25.08
Too close to call on our weighted score (GLM-4.5-Air 48, Codestral 25.08 47, Qwen3 235B-A22B Thinking 2507 42). The right pick depends on what you value most.
Z.ai (Zhipu)GLM-4.5-Air
48/100- ECI—
- Price$0.20 / $1.10
- Context131K
Alibaba (Qwen)Qwen3 235B-A22B Thinking 2507
42/100- ECI143.8
- Price$0.30 / $2.90
- Context262K
Mistral AICodestral 25.08
47/100- ECI—
- Price$0.30 / $0.90
- Context128K
Too close to call
It is close. Our weighted score puts them within a point (GLM-4.5-Air 48/100, Codestral 25.08 47/100, Qwen3 235B-A22B Thinking 2507 42/100), so choose by what matters most for your work: GLM-4.5-Air on price and Qwen3 235B-A22B Thinking 2507 for long inputs. 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 priceGLM-4.5-AirGLM-4.5-Air $0.425 · Codestral 25.08 $0.45 · Qwen3 235B-A22B Thinking 2507 $0.95 per 1M tokens (3:1 blend)
- Longest contextQwen3 235B-A22B Thinking 2507Qwen3 235B-A22B Thinking 2507 262,144 · GLM-4.5-Air 131,072 · Codestral 25.08 128,000 tokens
- Widest inputsSame inputsGLM-4.5-Air: Text · Qwen3 235B-A22B Thinking 2507: Text · Codestral 25.08: Text
- Self-hostingGLM-4.5-Air and Qwen3 235B-A22B Thinking 2507Publishes downloadable weights (Apache 2.0)
| Measure | Weight | GLM-4.5-Air | Qwen3 235B-A22B Thinking 2507 | Codestral 25.08 |
|---|---|---|---|---|
| Price | 50% | 65 | 49 | 64 |
| Inputs & features | 30% | 35 | 35 | 35 |
| Context window | 20% | 25 | 37 | 24 |
| Overall | 100% | 48/100 | 42/100 | 47/100 |
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 | GLM-4.5-AirZ.ai (Zhipu) | Qwen3 235B-A22B Thinking 2507Alibaba (Qwen) | Codestral 25.08Mistral AI |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | 143.8 | — |
| ECI rank | — | #93 of 165 | — |
| GPQA DiamondGraduate-level science questions | — | 80.1% | — |
| OTIS Mock AIME 2024–2025Competition mathematics | — | 86.7% | — |
| SimpleQA VerifiedShort factual questions | — | 40.4% | — |
| Price per million tokens | |||
| Input | $0.20 (best) | $0.30 | $0.30 |
| Output | $1.10 | $2.90 | $0.90 (best) |
| Cached input | $0.03 | — | $0.03 |
| Blended (3:1) | $0.425 (best) | $0.95 | $0.45 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Z.AI API | Median of 11 providers | Official Mistral API |
| Limits | |||
| Context window | 131,072 tokens | 262,144 tokens (best) | 128,000 tokens |
| Max output | 98,304 tokens (best) | 81,920 tokens | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | No | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yes | Yes | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | Yes |
| Availability | |||
| Weights | Open | OpenApache 2.0 | Proprietary |
| API model ID | glm-4.5-air | — | codestral-2508 |
| API providers | 19 (best) | 12 | 11 |
| Released | Jul 28, 2025 | Jul 25, 2025 | Jul 30, 2025 |
| Knowledge cutoff | 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.
- GLM-4.5-Air$4.20
- Qwen3 235B-A22B Thinking 2507$8.80
- Codestral 25.08$4.80
Which should you choose?
Which is better: GLM-4.5-Air, Qwen3 235B-A22B Thinking 2507 or Codestral 25.08?
It is close. Our weighted score puts them within a point (GLM-4.5-Air 48/100, Codestral 25.08 47/100, Qwen3 235B-A22B Thinking 2507 42/100), so choose by what matters most for your work: GLM-4.5-Air on price and Qwen3 235B-A22B Thinking 2507 for long inputs. 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, GLM-4.5-Air, Qwen3 235B-A22B Thinking 2507 or Codestral 25.08?
GLM-4.5-Air is cheaper at $0.20 input / $1.10 output per million tokens (official Z.AI API price). Codestral 25.08 costs $0.30 input / $0.90 output per million tokens (official Mistral API price); Qwen3 235B-A22B Thinking 2507 costs $0.30 input / $2.90 output per million tokens (median across 11 API providers). At a typical mix of three input tokens to one output token, that is $0.425 per million tokens for GLM-4.5-Air versus $0.45 for Codestral 25.08 (1.1× as much) and $0.95 for Qwen3 235B-A22B Thinking 2507 (2.2× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. GLM-4.5-Air has not been scored yet, Qwen3 235B-A22B Thinking 2507 has an ECI of 143.8 and Codestral 25.08 has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for GLM-4.5-Air, Qwen3 235B-A22B Thinking 2507 and Codestral 25.08 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 235B-A22B Thinking 2507 has the largest context window at 262,144 tokens, against 131,072 for GLM-4.5-Air and 128,000 for Codestral 25.08. Maximum output per response: GLM-4.5-Air up to 98,304, Qwen3 235B-A22B Thinking 2507 up to 81,920, Codestral 25.08 up to 8,192 tokens.
Which can read images, PDFs, audio or video?
GLM-4.5-Air accepts text; Qwen3 235B-A22B Thinking 2507 accepts text; Codestral 25.08 accepts text. They handle the same number of input types.
Are any of these open source?
GLM-4.5-Air and Qwen3 235B-A22B Thinking 2507 publishes its weights (Apache 2.0) and can be self-hosted; Codestral 25.08 is proprietary.
Which is newer?
Codestral 25.08 is the newest, released Jul 30, 2025. GLM-4.5-Air came out Jul 28, 2025; Qwen3 235B-A22B Thinking 2507 came out Jul 25, 2025. Knowledge cutoff: GLM-4.5-Air 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 Nov 2, 2026.
Related comparisons
- GLM-4.5-Air vs Llama 3.3 Nemotron Super 49B v1.5NVIDIA
- GLM-4.5-Air vs Qwen3 Omni 30B A3B ThinkingAlibaba (Qwen)
- GLM-4.5-Air vs ERNIE 4.5 300B A47BBaidu
- Qwen3 235B-A22B Thinking 2507 vs GLM-4.7Z.ai (Zhipu) · ECI 143.7
- Qwen3 235B-A22B Thinking 2507 vs GPT-5 MiniOpenAI · ECI 145.4
- Qwen3 235B-A22B Thinking 2507 vs DeepSeek R1 0528DeepSeek · ECI 141.2
