GLM-5.2 vs Kimi K2.6 vs Qwen3.8 Max
Qwen3.8 Max comes out ahead, 69 to 65 and 61 on our weighted score, though Kimi K2.6 is 43% cheaper per token.
Z.ai (Zhipu)GLM-5.2
61/100- ECI151.8
- Price$1.40 / $4.40
- Context1M
Moonshot AIKimi K2.6
65/100- ECI151.1
- Price$0.95 / $4.00
- Context262K
- Our pick
Alibaba (Qwen)Qwen3.8 Max
69/100- ECI156.6
- Price$2.00 / $6.00
- Context1M
Qwen3.8 Max is our pick
Qwen3.8 Max is the better all-round choice, scoring 69/100 against Kimi K2.6 (65) and GLM-5.2 (61). It leads on capability and inputs & features. Kimi K2.6 wins on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityQwen3.8 MaxCapabilities Index (ECI): Qwen3.8 Max 156.6 · GLM-5.2 151.8 · Kimi K2.6 151.1
- Lowest priceKimi K2.6Kimi K2.6 $1.71 · GLM-5.2 $2.15 · Qwen3.8 Max $3.00 per 1M tokens (3:1 blend)
- Longest contextGLM-5.2 and Qwen3.8 MaxGLM-5.2 1,000,000 · Qwen3.8 Max 1,000,000 · Kimi K2.6 262,144 tokens
- Widest inputsQwen3.8 MaxGLM-5.2: Text · Kimi K2.6: Text, Images, Video · Qwen3.8 Max: Text, Images, PDFs, Video
- Self-hostingGLM-5.2 and Kimi K2.6Publishes downloadable weights
| Measure | Weight | GLM-5.2 | Kimi K2.6 | Qwen3.8 Max |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 80 | 79 | 86 |
| Price | 25% | 34 | 39 | 27 |
| Inputs & features | 15% | 45 | 80 | 90 |
| Context window | 10% | 60 | 37 | 60 |
| Overall | 100% | 61/100 | 65/100 | 69/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | GLM-5.2Z.ai (Zhipu) | Kimi K2.6Moonshot AI | Qwen3.8 MaxAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 151.8 | 151.1 | 156.6 (best) |
| ECI rank | #44 of 148 | #45 of 148 | #20 of 148 (best) |
| GPQA DiamondGraduate-level science questions | 91.9% | 90.8% | 92.7% (best) |
| FrontierMath Tiers 1–3Research-level mathematics | 59.2% | 57.2% | 74.7% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | 86.4% | 96.1% | 99.4% (best) |
| SWE-bench VerifiedFixing real GitHub issues | 78.7% (best) | 76.7% | — |
| SimpleQA VerifiedShort factual questions | 34.2% | 34.9% | 45.8% (best) |
| Price per million tokens | |||
| Input | $1.40 | $0.95 (best) | $2.00 |
| Output | $4.40 | $4.00 (best) | $6.00 |
| Cached input | $0.26 | $0.16 (best) | $0.25 |
| Blended (3:1) | $2.15 | $1.71 (best) | $3.00 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Z.AI API | Official Moonshot AI API | Official Alibaba API |
| Limits | |||
| Context window | 1,000,000 tokens (best) | 262,144 tokens | 1,000,000 tokens (best) |
| Max output | 131,072 tokens | 262,144 tokens (best) | 131,072 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | Yes |
| Audio | No | No | No |
| Video | No | Yes | Yes |
| Reasoning | Yeshigh · max | Yes | Yeslow · medium · xhigh |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | Yes | Yes |
| Availability | |||
| Weights | Open | Open | Proprietary |
| API model ID | glm-5.2 | kimi-k2.6 | qwen3.8-max |
| API providers | 80 (best) | 46 | 25 |
| Released | Jun 13, 2026 | Apr 21, 2026 | Aug 3, 2026 |
| Knowledge cutoff | — | Jan 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-5.2$22.80
- Kimi K2.6$17.50
- Qwen3.8 Max$32.00
Which should you choose?
Which is better: GLM-5.2, Kimi K2.6 or Qwen3.8 Max?
Qwen3.8 Max is the better all-round choice, scoring 69/100 against Kimi K2.6 (65) and GLM-5.2 (61). It leads on capability and inputs & features. Kimi K2.6 wins on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, GLM-5.2, Kimi K2.6 or Qwen3.8 Max?
Kimi K2.6 is cheaper at $0.95 input / $4.00 output per million tokens (official Moonshot AI API price). GLM-5.2 costs $1.40 input / $4.40 output per million tokens (official Z.AI API price); Qwen3.8 Max costs $2.00 input / $6.00 output per million tokens (official Alibaba API price). At a typical mix of three input tokens to one output token, that is $1.71 per million tokens for Kimi K2.6 versus $2.15 for GLM-5.2 (1.3× as much) and $3.00 for Qwen3.8 Max (1.8× as much).
Which scores higher on benchmarks?
Qwen3.8 Max scores higher on the Capabilities Index (ECI): Qwen3.8 Max 156.6 (#20 of 148), GLM-5.2 151.8 (#44 of 148) and Kimi K2.6 151.1 (#45 of 148). Their confidence ranges do not overlap (154.5–158.8 vs 149.8–154.0), so the gap is a real one. On individual benchmarks: GPQA Diamond — Qwen3.8 Max 92.7%, GLM-5.2 91.9%, Kimi K2.6 90.8%; FrontierMath Tiers 1–3 — Qwen3.8 Max 74.7%, GLM-5.2 59.2%, Kimi K2.6 57.2%; OTIS Mock AIME 2024–2025 — Qwen3.8 Max 99.4%, Kimi K2.6 96.1%, GLM-5.2 86.4%; SimpleQA Verified — Qwen3.8 Max 45.8%, Kimi K2.6 34.9%, GLM-5.2 34.2%.
Which is better for coding?
There are no published SWE-bench Verified results for Qwen3.8 Max yet, so there is no like-for-like coding score. On overall capability, Qwen3.8 Max 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?
GLM-5.2 and Qwen3.8 Max have the largest context windows (1,000,000 and 1,000,000 tokens), against 262,144 for Kimi K2.6. Maximum output per response: GLM-5.2 up to 131,072, Kimi K2.6 up to 262,144, Qwen3.8 Max up to 131,072 tokens.
Which can read images, PDFs, audio or video?
GLM-5.2 accepts text; Kimi K2.6 accepts text, images and video; Qwen3.8 Max accepts text, images, PDFs and video. Qwen3.8 Max handles the widest range of inputs.
Are any of these open source?
GLM-5.2 and Kimi K2.6 publishes its weights and can be self-hosted; Qwen3.8 Max is proprietary.
Which is newer?
Qwen3.8 Max is the newest, released Aug 3, 2026. GLM-5.2 came out Jun 13, 2026; Kimi K2.6 came out Apr 21, 2026. Knowledge cutoff: Kimi K2.6 Jan 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.
