Claude Sonnet 5 vs Kimi K3 vs Qwen3.8 Max
Qwen3.8 Max comes out ahead, 69 to 66 and 65 on our weighted score, and it is the cheaper option too.
AnthropicClaude Sonnet 5
66/100- ECI156.3
- Price$2.00 / $10.00
- Context1M
Moonshot AIKimi K3
65/100- ECI157.6
- Price$3.00 / $15.00
- Context1.05M
- 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 Claude Sonnet 5 (66) and Kimi K3 (65). It leads on price and inputs & features. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityKimi K3Capabilities Index (ECI): Kimi K3 157.6 · Qwen3.8 Max 156.6 · Claude Sonnet 5 156.3
- Lowest priceQwen3.8 MaxQwen3.8 Max $3.00 · Claude Sonnet 5 $4.00 · Kimi K3 $6.00 per 1M tokens (3:1 blend)
- Longest contextKimi K3Kimi K3 1,048,576 · Claude Sonnet 5 1,000,000 · Qwen3.8 Max 1,000,000 tokens
- Widest inputsQwen3.8 MaxClaude Sonnet 5: Text, Images, PDFs · Kimi K3: Text, Images, Video · Qwen3.8 Max: Text, Images, PDFs, Video
- Self-hostingKimi K3Publishes downloadable weights
| Measure | Weight | Claude Sonnet 5 | Kimi K3 | Qwen3.8 Max |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 86 | 88 | 86 |
| Price | 25% | 21 | 13 | 27 |
| Inputs & features | 15% | 80 | 80 | 90 |
| Context window | 10% | 60 | 61 | 60 |
| Overall | 100% | 66/100 | 65/100 | 69/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Claude Sonnet 5Anthropic | Kimi K3Moonshot AI | Qwen3.8 MaxAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 156.3 | 157.6 (best) | 156.6 |
| ECI rank | #23 of 148 | #13 of 148 (best) | #20 of 148 |
| GPQA DiamondGraduate-level science questions | 90.5% | 93.1% (best) | 92.7% |
| FrontierMath Tiers 1–3Research-level mathematics | 65.6% | 72.2% | 74.7% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | 94.7% | 97.2% | 99.4% (best) |
| SimpleQA VerifiedShort factual questions | 33.7% | 50.6% (best) | 45.8% |
| Price per million tokens | |||
| Input | $2.00 (best) | $3.00 | $2.00 (best) |
| Output | $10.00 | $15.00 | $6.00 (best) |
| Cached input | $0.20 (best) | $0.30 | $0.25 |
| Blended (3:1) | $4.00 | $6.00 | $3.00 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Anthropic API | Official Moonshot AI API | Official Alibaba API |
| Limits | |||
| Context window | 1,000,000 tokens | 1,048,576 tokens (best) | 1,000,000 tokens |
| Max output | 128,000 tokens | 131,072 tokens (best) | 131,072 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | Yes |
| PDFs | Yes | No | Yes |
| Audio | No | No | No |
| Video | No | Yes | Yes |
| Reasoning | Yeslow · medium · high · xhigh · max | Yeslow · high · max | Yeslow · medium · xhigh |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | Yes | Yes |
| Availability | |||
| Weights | Proprietary | Open | Proprietary |
| API model ID | claude-sonnet-5 | kimi-k3 | qwen3.8-max |
| API providers | 38 | 68 (best) | 25 |
| Released | Jun 30, 2026 | Jul 16, 2026 | Aug 3, 2026 |
| Knowledge cutoff | Jan 31, 2026 | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Claude Sonnet 5$40.00
- Kimi K3$60.00
- Qwen3.8 Max$32.00
Which should you choose?
Which is better: Claude Sonnet 5, Kimi K3 or Qwen3.8 Max?
Qwen3.8 Max is the better all-round choice, scoring 69/100 against Claude Sonnet 5 (66) and Kimi K3 (65). It leads on price and inputs & features. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, Claude Sonnet 5, Kimi K3 or Qwen3.8 Max?
Qwen3.8 Max is cheaper at $2.00 input / $6.00 output per million tokens (official Alibaba API price). Claude Sonnet 5 costs $2.00 input / $10.00 output per million tokens (official Anthropic API price); Kimi K3 costs $3.00 input / $15.00 output per million tokens (official Moonshot AI API price). At a typical mix of three input tokens to one output token, that is $3.00 per million tokens for Qwen3.8 Max versus $4.00 for Claude Sonnet 5 (1.3× as much) and $6.00 for Kimi K3 (2× as much).
Which scores higher on benchmarks?
Kimi K3 scores higher on the Capabilities Index (ECI): Kimi K3 157.6 (#13 of 148), Qwen3.8 Max 156.6 (#20 of 148) and Claude Sonnet 5 156.3 (#23 of 148). The confidence ranges of the top two overlap (154.9–160.4 vs 154.5–158.8), so treat the gap as small. On individual benchmarks: GPQA Diamond — Kimi K3 93.1%, Qwen3.8 Max 92.7%, Claude Sonnet 5 90.5%; FrontierMath Tiers 1–3 — Qwen3.8 Max 74.7%, Kimi K3 72.2%, Claude Sonnet 5 65.6%; OTIS Mock AIME 2024–2025 — Qwen3.8 Max 99.4%, Kimi K3 97.2%, Claude Sonnet 5 94.7%; SimpleQA Verified — Kimi K3 50.6%, Qwen3.8 Max 45.8%, Claude Sonnet 5 33.7%.
Which is better for coding?
There are no published SWE-bench Verified results for Claude Sonnet 5, Kimi K3 and Qwen3.8 Max yet, so there is no like-for-like coding score. On overall capability, Kimi K3 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 K3 has the largest context window at 1,048,576 tokens, against 1,000,000 for Claude Sonnet 5 and 1,000,000 for Qwen3.8 Max. Maximum output per response: Claude Sonnet 5 up to 128,000, Kimi K3 up to 131,072, Qwen3.8 Max up to 131,072 tokens.
Which can read images, PDFs, audio or video?
Claude Sonnet 5 accepts text, images and PDFs; Kimi K3 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?
Kimi K3 publishes its weights and can be self-hosted; Claude Sonnet 5 and Qwen3.8 Max is proprietary.
Which is newer?
Qwen3.8 Max is the newest, released Aug 3, 2026. Kimi K3 came out Jul 16, 2026; Claude Sonnet 5 came out Jun 30, 2026. Knowledge cutoff: Claude Sonnet 5 Jan 31, 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.
