Kimi K2 Instruct vs Nemotron 3 Ultra 550B A55B vs Qwen3.5 Plus
Qwen3.5 Plus comes out ahead, 74 to 67 and 55 on our weighted score, and it is the cheaper option too.
Price in rating: Excluded · Include price
Moonshot AI
Kimi K2 Instruct
55/100- ECI141.3
- Price$0.57 / $2.30
- Context131K
NVIDIA
Nemotron 3 Ultra 550B A55B
67/100- ECI147.0
- Price$0.55 / $2.40
- Context1M
- Our pick
Alibaba (Qwen)
Qwen3.5 Plus
74/100- ECI147.3
- Price$0.40 / $2.40
- Context1M
Qwen3.5 Plus is our pick
Qwen3.5 Plus is the better all-round choice, scoring 74/100 against Nemotron 3 Ultra 550B A55B (67) and Kimi K2 Instruct (55). It leads on inputs & features. The score weighs capability 67%, inputs & features 20%, context window 13%.
- CapabilityQwen3.5 PlusCapabilities Index (ECI): Qwen3.5 Plus 147.3 · Nemotron 3 Ultra 550B A55B 147.0 · Kimi K2 Instruct 141.3
- Lowest priceQwen3.5 PlusQwen3.5 Plus $0.90 · Kimi K2 Instruct $1.00 · Nemotron 3 Ultra 550B A55B $1.01 per 1M tokens (3:1 blend)
- Longest contextNemotron 3 Ultra 550B A55B and Qwen3.5 PlusNemotron 3 Ultra 550B A55B 1,000,000 · Qwen3.5 Plus 1,000,000 · Kimi K2 Instruct 131,072 tokens
- Widest inputsQwen3.5 PlusKimi K2 Instruct: Text · Nemotron 3 Ultra 550B A55B: Text · Qwen3.5 Plus: Text, Images, Video
- Self-hostingKimi K2 Instruct and Nemotron 3 Ultra 550B A55BPublishes downloadable weights (Modified MIT)
| Measure | Weight | Kimi K2 Instruct | Nemotron 3 Ultra 550B A55B | Qwen3.5 Plus |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 67% | 69 | 74 | 75 |
| Inputs & features | 20% | 25 | 45 | 80 |
| Context window | 13% | 32 | 64 | 64 |
| Overall | 100% | 55/100 | 67/100 | 74/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 141.3 | 147.0 | 147.3 (best) |
| ECI rank | #108 of 158 | #82 of 158 | #80 of 158 (best) |
| GPQA DiamondGraduate-level science questions | — | 85.4% (best) | 84.9% |
| OTIS Mock AIME 2024–2025Competition mathematics | — | 86.7% | 86.7% |
| SimpleQA VerifiedShort factual questions | — | — | 25.4% |
| Price per million tokens | |||
| Input | $0.57 | $0.55 | $0.40 (best) |
| Output | $2.30 (best) | $2.40 | $2.40 |
| Cached input | — | — | — |
| Blended (3:1) | $1.00 | $1.01 | $0.90 (best) |
| Long-context rate | Same rate | Same rate | Over 256K: $0.50 / $3.00 |
| Price source | Median of 11 providers | Median of 20 providers | Official Alibaba API |
| Limits | |||
| Context window | 131,072 tokens | 1,000,000 tokens (best) | 1,000,000 tokens (best) |
| Max output | 32,768 tokens | 128,000 tokens (best) | 65,536 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | No | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | No | Yesmedium · high | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | Yes |
| Availability | |||
| Weights | OpenModified MIT | Open | Proprietary |
| API model ID | — | nvidia/nemotron-3-ultra-550b-a55b | qwen3.5-plus |
| API providers | 11 | 26 (best) | 14 |
| Released | Jul 11, 2025 | Jun 4, 2026 | Feb 16, 2026 |
| Knowledge cutoff | — | Sep 2025 | 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.
Kimi K2 Instruct$10.30
Nemotron 3 Ultra 550B A55B$10.30
Qwen3.5 Plus$8.80
Which should you choose?
Which is better: Kimi K2 Instruct, Nemotron 3 Ultra 550B A55B or Qwen3.5 Plus?
Qwen3.5 Plus is the better all-round choice, scoring 74/100 against Nemotron 3 Ultra 550B A55B (67) and Kimi K2 Instruct (55). It leads on inputs & features. The score weighs capability 67%, inputs & features 20%, context window 13%.
Which is cheaper, Kimi K2 Instruct, Nemotron 3 Ultra 550B A55B or Qwen3.5 Plus?
Qwen3.5 Plus is cheaper at $0.40 input / $2.40 output per million tokens (official Alibaba API price). Kimi K2 Instruct costs $0.57 input / $2.30 output per million tokens (median across 11 API providers); Nemotron 3 Ultra 550B A55B costs $0.55 input / $2.40 output per million tokens (median across 20 API providers; free on Nvidia). At a typical mix of three input tokens to one output token, that is $0.90 per million tokens for Qwen3.5 Plus versus $1.00 for Kimi K2 Instruct (1.1× as much) and $1.01 for Nemotron 3 Ultra 550B A55B (1.1× as much).
Which scores higher on benchmarks?
Qwen3.5 Plus scores higher on the Capabilities Index (ECI): Qwen3.5 Plus 147.3 (#80 of 158), Nemotron 3 Ultra 550B A55B 147.0 (#82 of 158) and Kimi K2 Instruct 141.3 (#108 of 158). The confidence ranges of the top two overlap (145.4–149.2 vs 145.0–149.1), so treat the gap as small.
Which is better for coding?
There are no published SWE-bench Verified results for Kimi K2 Instruct, Nemotron 3 Ultra 550B A55B and Qwen3.5 Plus yet, so there is no like-for-like coding score. On overall capability, Qwen3.5 Plus 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?
Nemotron 3 Ultra 550B A55B and Qwen3.5 Plus have the largest context windows (1,000,000 and 1,000,000 tokens), against 131,072 for Kimi K2 Instruct. Maximum output per response: Kimi K2 Instruct up to 32,768, Nemotron 3 Ultra 550B A55B up to 128,000, Qwen3.5 Plus up to 65,536 tokens.
Which can read images, PDFs, audio or video?
Kimi K2 Instruct accepts text; Nemotron 3 Ultra 550B A55B accepts text; Qwen3.5 Plus accepts text, images and video. Qwen3.5 Plus handles the widest range of inputs.
Are any of these open source?
Kimi K2 Instruct and Nemotron 3 Ultra 550B A55B publishes its weights (Modified MIT) and can be self-hosted; Qwen3.5 Plus is proprietary.
Which is newer?
Nemotron 3 Ultra 550B A55B is the newest, released Jun 4, 2026. Qwen3.5 Plus came out Feb 16, 2026; Kimi K2 Instruct came out Jul 11, 2025. Knowledge cutoff: Nemotron 3 Ultra 550B A55B Sep 2025, Qwen3.5 Plus Apr 2025.
How do you decide the winner?
This rating uses capability (67%, independent benchmark results); inputs & features (20%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (13%, 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 18, 2026.