Qwen3.5 397B-A17B vs Qwen3.8 27B vs GLM-5
Too close to call on our weighted score (Qwen3.8 27B 67, Qwen3.5 397B-A17B 65, GLM-5 55). The right pick depends on what you value most.
Alibaba (Qwen)
Qwen3.5 397B-A17B
65/100- ECI146.7
- Price$0.60 / $3.60
- Context262K
Alibaba (Qwen)
Qwen3.8 27B
67/100- ECI149.4
- Price$0.40 / $2.50
- Context262K
Z.ai (Zhipu)
GLM-5
55/100- ECI145.8
- Price$1.00 / $3.20
- Context205K
Too close to call
It is close. Our weighted score puts them within 2 points (Qwen3.8 27B 67/100, Qwen3.5 397B-A17B 65/100, GLM-5 55/100), so choose by what matters most for your work: Qwen3.8 27B for raw capability. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityQwen3.8 27BCapabilities Index (ECI): Qwen3.8 27B 149.4 · Qwen3.5 397B-A17B 146.7 · GLM-5 145.8
- Lowest priceQwen3.8 27BQwen3.8 27B $0.925 · Qwen3.5 397B-A17B $1.35 · GLM-5 $1.55 per 1M tokens (3:1 blend)
- Longest contextQwen3.5 397B-A17B and Qwen3.8 27BQwen3.5 397B-A17B 262,144 · Qwen3.8 27B 262,144 · GLM-5 204,800 tokens
- Widest inputsQwen3.5 397B-A17BQwen3.5 397B-A17B: Text, Images, Audio, Video · Qwen3.8 27B: Text, Images, Video · GLM-5: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Qwen3.5 397B-A17B | Qwen3.8 27B | GLM-5 |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 74 | 77 | 73 |
| Price | 25% | 44 | 51 | 41 |
| Inputs & features | 15% | 90 | 80 | 35 |
| Context window | 10% | 37 | 37 | 32 |
| Overall | 100% | 65/100 | 67/100 | 55/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 146.7 | 149.4 (best) | 145.8 |
| ECI rank | #67 of 148 | #53 of 148 (best) | #74 of 148 |
| GPQA DiamondGraduate-level science questions | 86.4% | — | 87.8% (best) |
| FrontierMath Tiers 1–3Research-level mathematics | 31.2% | — | — |
| OTIS Mock AIME 2024–2025Competition mathematics | 88.9% (best) | — | 80.0% |
| SWE-bench VerifiedFixing real GitHub issues | — | — | 72.1% |
| Price per million tokens | |||
| Input | $0.60 | $0.40 (best) | $1.00 |
| Output | $3.60 | $2.50 (best) | $3.20 |
| Cached input | — | — | $0.20 |
| Blended (3:1) | $1.35 | $0.925 (best) | $1.55 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Alibaba API | Median of 39 providers | Official Z.AI API |
| Limits | |||
| Context window | 262,144 tokens (best) | 262,144 tokens (best) | 204,800 tokens |
| Max output | 65,536 tokens | 32,768 tokens | 131,072 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | No |
| PDFs | No | No | No |
| Audio | Yes | No | No |
| Video | Yes | Yes | No |
| Reasoning | Yes | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | Yes | No |
| Availability | |||
| Weights | Open | Open | Open |
| API model ID | qwen3.5-397b-a17b | — | glm-5 |
| API providers | 23 | 41 (best) | 27 |
| Released | Feb 15, 2026 | Aug 14, 2026 | Feb 12, 2026 |
| Knowledge cutoff | — | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
Qwen3.5 397B-A17B$13.20
Qwen3.8 27B$9.00
GLM-5$16.40
Which should you choose?
Which is better: Qwen3.5 397B-A17B, Qwen3.8 27B or GLM-5?
It is close. Our weighted score puts them within 2 points (Qwen3.8 27B 67/100, Qwen3.5 397B-A17B 65/100, GLM-5 55/100), so choose by what matters most for your work: Qwen3.8 27B for raw capability. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, Qwen3.5 397B-A17B, Qwen3.8 27B or GLM-5?
Qwen3.8 27B is cheaper at $0.40 input / $2.50 output per million tokens (median across 39 API providers). Qwen3.5 397B-A17B costs $0.60 input / $3.60 output per million tokens (official Alibaba API price); GLM-5 costs $1.00 input / $3.20 output per million tokens (official Z.AI API price). At a typical mix of three input tokens to one output token, that is $0.925 per million tokens for Qwen3.8 27B versus $1.35 for Qwen3.5 397B-A17B (1.5× as much) and $1.55 for GLM-5 (1.7× as much).
Which scores higher on benchmarks?
Qwen3.8 27B scores higher on the Capabilities Index (ECI): Qwen3.8 27B 149.4 (#53 of 148), Qwen3.5 397B-A17B 146.7 (#67 of 148) and GLM-5 145.8 (#74 of 148). The confidence ranges of the top two overlap (147.5–151.6 vs 144.8–148.2), so treat the gap as small.
Which is better for coding?
There are no published SWE-bench Verified results for Qwen3.5 397B-A17B and Qwen3.8 27B yet, so there is no like-for-like coding score. On overall capability, Qwen3.8 27B 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?
Qwen3.5 397B-A17B and Qwen3.8 27B have the largest context windows (262,144 and 262,144 tokens), against 204,800 for GLM-5. Maximum output per response: Qwen3.5 397B-A17B up to 65,536, Qwen3.8 27B up to 32,768, GLM-5 up to 131,072 tokens.
Which can read images, PDFs, audio or video?
Qwen3.5 397B-A17B accepts text, images, audio and video; Qwen3.8 27B accepts text, images and video; GLM-5 accepts text. Qwen3.5 397B-A17B 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?
Qwen3.8 27B is the newest, released Aug 14, 2026. Qwen3.5 397B-A17B came out Feb 15, 2026; GLM-5 came out Feb 12, 2026.
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.