Sakana Namazu vs Hy4 preview vs Kimi K2.8 Preview
Too close to call on our weighted score (Kimi K2.8 Preview 51, Sakana Namazu 50, Hy4 preview 44). The right pick depends on what you value most.
Sakana AISakana Namazu
50/100- ECI—
- Price$0.95 / $4.00
- Context262K
TencentHy4 preview
44/100- ECI—
- Price$0.834 / $2.50
- Context1.02M
Moonshot AIKimi K2.8 Preview
51/100- ECI—
- Price$1.00 / $4.00
- Context1.05M
Too close to call
It is close. Our weighted score puts them within 2 points (Kimi K2.8 Preview 51/100, Sakana Namazu 50/100, Hy4 preview 44/100), so choose by what matters most for your work: Hy4 preview on price and Kimi K2.8 Preview 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 priceHy4 previewHy4 preview $1.25 · Sakana Namazu $1.71 · Kimi K2.8 Preview $1.75 per 1M tokens (3:1 blend)
- Longest contextKimi K2.8 PreviewKimi K2.8 Preview 1,048,576 · Hy4 preview 1,024,000 · Sakana Namazu 262,144 tokens
- Widest inputsSakana Namazu and Kimi K2.8 PreviewSakana Namazu: Text, Images, PDFs · Hy4 preview: Text · Kimi K2.8 Preview: Text, Images, Video
- Self-hostingHy4 previewPublishes downloadable weights
| Measure | Weight | Sakana Namazu | Hy4 preview | Kimi K2.8 Preview |
|---|---|---|---|---|
| Price | 50% | 37 | 43 | 37 |
| Inputs & features | 30% | 80 | 35 | 70 |
| Context window | 20% | 37 | 60 | 61 |
| Overall | 100% | 50/100 | 44/100 | 51/100 |
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 | Sakana NamazuSakana AI | Hy4 previewTencent | Kimi K2.8 PreviewMoonshot AI |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.95 | $0.834 (best) | $1.00 |
| Output | $4.00 | $2.50 (best) | $4.00 |
| Cached input | $0.15 | $0.042 (best) | — |
| Blended (3:1) | $1.71 | $1.25 (best) | $1.75 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Sakana AI API | Official Tencent TokenHub API | Median of 2 providers |
| Limits | |||
| Context window | 262,144 tokens | 1,024,000 tokens | 1,048,576 tokens (best) |
| Max output | 65,536 tokens (best) | 64,000 tokens | — |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | No | Yes |
| PDFs | Yes | No | No |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | Yes | Yeshigh | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | No | No |
| Availability | |||
| Weights | Proprietary | Open | Proprietary |
| API model ID | sakana-namazu | hy4-preview | — |
| API providers | 5 | 17 (best) | 2 |
| Released | Aug 3, 2026 | Aug 28, 2026 | Sep 11, 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.
- Sakana Namazu$17.50
- Hy4 preview$13.34
- Kimi K2.8 Preview$18.00
Which should you choose?
Which is better: Sakana Namazu, Hy4 preview or Kimi K2.8 Preview?
It is close. Our weighted score puts them within 2 points (Kimi K2.8 Preview 51/100, Sakana Namazu 50/100, Hy4 preview 44/100), so choose by what matters most for your work: Hy4 preview on price and Kimi K2.8 Preview 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, Sakana Namazu, Hy4 preview or Kimi K2.8 Preview?
Hy4 preview is cheaper at $0.834 input / $2.50 output per million tokens (official Tencent TokenHub API price). Sakana Namazu costs $0.95 input / $4.00 output per million tokens (official Sakana AI API price); Kimi K2.8 Preview costs $1.00 input / $4.00 output per million tokens (median across 2 API providers). At a typical mix of three input tokens to one output token, that is $1.25 per million tokens for Hy4 preview versus $1.71 for Sakana Namazu (1.4× as much) and $1.75 for Kimi K2.8 Preview (1.4× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Sakana Namazu has not been scored yet, Hy4 preview has not been scored yet and Kimi K2.8 Preview has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Sakana Namazu, Hy4 preview and Kimi K2.8 Preview 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?
Kimi K2.8 Preview has the largest context window at 1,048,576 tokens, against 1,024,000 for Hy4 preview and 262,144 for Sakana Namazu. Maximum output per response: Sakana Namazu up to 65,536, Hy4 preview up to 64,000 tokens.
Which can read images, PDFs, audio or video?
Sakana Namazu accepts text, images and PDFs; Hy4 preview accepts text; Kimi K2.8 Preview accepts text, images and video. Sakana Namazu handles the widest range of inputs.
Are any of these open source?
Hy4 preview publishes its weights and can be self-hosted; Sakana Namazu and Kimi K2.8 Preview is proprietary.
Which is newer?
Kimi K2.8 Preview is the newest, released Sep 11, 2026. Hy4 preview came out Aug 28, 2026; Sakana Namazu came out Aug 3, 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 Nov 2, 2026.
