Nebius: Models Intelligence, Performance & Price
Analysis of Nebius's models across key metrics including quality, price, output speed, latency, context window & more. This analysis is intended to support you in choosing the best model provided by Nebius for your use-case.
Most Intelligent
Intelligence index
Total 30 models
Fastest
Output speed
Total 30 models
Lowest Price
Blended price (per 1M tokens)
Total 30 models
Nebius offers 30 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across models.
- For intelligence, the top models on Nebius are Kimi K3 (max) (57), GLM-5.2 (max) (FP4) (51), and MiniMax-M3 (FP8) (44).
- For output speed, the fastest models are GLM-5.2 (max) (FP4) (353 t/s), Nemotron 3 Ultra (326 t/s), and NVIDIA Nemotron 3 Super (323 t/s).
- For latency, Qwen3 30B A3B 2507 (1.12s), GLM-5.2 (FP4) (1.39s), and Hermes 4 70B (FP8) (1.39s) offer the lowest time to first answer token.
- For pricing, Nemotron 3 Nano Omni 30B A3B Reasoning (FP8) ($0.08), NVIDIA Nemotron 3 Nano ($0.08), and Qwen3 32B Base ($0.12) offer the lowest blended prices per 1M tokens.
- For context window size, MiniMax-M3 (FP8) (1M), Kimi K3 (max) (1M), and GLM-5.2 (max) (FP4) (432k) support the largest context windows on Nebius.
Highlights
Intelligence Evaluations
Artificial Analysis Intelligence Index
Intelligence Evaluations
Agentic real-world work tasks, (Elo-500)/2000
Agentic tool use
Agentic coding & terminal use
Coding
Reasoning & knowledge
Scientific reasoning
Physics reasoning
Knowledge
1 - hallucination rate
Long context reasoning
Agentic knowledge work, Elo
Agentic SaaS workflows
Legal agentic work, criterion pass rate
Agentic business operations
Instruction following
Long-horizon agentic tasks
Kubernetes incident root-cause analysis
Visual reasoning
Intelligence Index vs. Price
Context Window
Context Window
Pricing
Intelligence Index vs. Price
Performance Summary
Output Speed vs. Price
Speed
Measured by Output Speed (tokens per second)
Output Speed
Latency
Measured by Time (seconds) to First Token
Latency: Time To First Answer Token
End-to-End Response Time
Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed
End-to-End Response Time vs. Price
Further Analysis | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
Kimi K3 (max) | 1.05M | Open | 57 | $3.18 | 83 | 1.69 | 31.71 | 24.02 | |||
GLM-5.2 (max) (FP4) | 432k | Open | 51 | $1.06 | 353 | 1.13 | 8.21 | 5.67 | |||
MiniMax-M3 (FP8) | 1.05M | Open | 44 | $0.45 | 183 | 1.95 | 15.59 | 10.91 | |||
Kimi K2.6 | 262k | Open | 44 | $0.76 | 246 | 1.92 | 22.05 | 18.10 | |||
Kimi K2.7 Code (FP4) | 256k | Open | 42 | $0.64 | 231 | 1.82 | 13.64 | 9.66 | |||
GLM-5.1 (FP8, Base) | 200k | Open | 40 | $0.58 | 33 | 1.79 | 130.87 | 114.04 | |||
Nemotron 3 Ultra | 256k | Open | 38 | $0.59 | 326 | 2.28 | 10.79 | 6.98 | |||
GLM-5.1 (FP8, Base) | 200k | Open | 35* | -- | 30 | 1.81 | 18.42 | -- | |||
Kimi K2.6 | 262k | Open | 35* | -- | 190 | 1.96 | 4.60 | -- | |||
GLM-5.2 (FP4) | 432k | Open | 34 | -- | 265 | 1.39 | 3.27 | -- | |||
Qwen3.5 397B A17B (Base, FP4) | 262k | Open | 34 | $0.36 | 89 | 2.02 | 43.45 | 35.81 | |||
MiniMax-M2.5 (FP4) | 196k | Open | 34* | -- | 101 | 1.86 | 26.51 | 19.72 | |||
Qwen3.5 397B A17B (Base, FP4) | 262k | Open | 32* | -- | 91 | 2.06 | 7.57 | -- | |||
NVIDIA Nemotron 3 Super | 256k | Open | 25 | $0.34 | 323 | 1.83 | 9.57 | 6.20 | |||
 gpt-oss-120b (high) (Base) | 128k | Open | 24 | $0.08 | 242 | 1.02 | 11.37 | 8.27 | |||
Qwen3 235B 2507 | 262k | Open | 18* | -- | 44 | 1.70 | 12.95 | -- | |||
Qwen3 Next 80B A3B (FP8) | 262k | Open | 17 | $0.04 | 103 | 1.03 | 25.32 | 19.43 | |||
Nemotron 3 Nano Omni 30B A3B Reasoning (FP8) | 65.5k | Open | 15* | -- | 300 | 0.99 | 9.32 | 6.67 | |||
gpt-oss-120b (low) Base | 128k | Open | 15 | $0.02 | 243 | 1.02 | 11.30 | 8.23 | |||
NVIDIA Nemotron 3 Nano | 262k | Open | 14 | $0.03 | 320 | 1.03 | 8.85 | 6.25 | |||
Qwen3 32B Base | 32.8k | Open | 12 | -- | 29 | 1.89 | 86.77 | 67.91 | |||
 Hermes 4 70B (FP8) | 128k | Open | 10* | -- | 85 | 1.39 | 30.92 | 23.62 | |||
 Llama 3.3 70B Base | 128k | Open | 9 | $0.02 | 13 | 3.04 | 40.43 | -- | |||
Llama Nemotron Ultra Base | 131k | Open | 9* | -- | 51 | 2.39 | 51.28 | 39.11 | |||
Qwen3 30B A3B 2507 | 262k | Open | 9* | -- | 53 | 1.12 | 10.60 | -- | |||
Hermes 4 405B (FP8) | 128k | Open | 9* | -- | 38 | 2.38 | 68.41 | 52.82 | |||
Hermes 4 405B (FP8) | 128k | Open | 9* | -- | 33 | 2.37 | 17.30 | -- | |||
Qwen3 32B Base | 32.8k | Open | 9* | -- | 30 | 1.87 | 18.50 | -- | |||
 Gemma 3 27B (FP8) | 110k | Open | 7 | $0.09 | 23 | 3.28 | 25.34 | -- | |||
Hermes 4 70B (FP8) | 128k | Open | 7* | -- | 94 | 1.39 | 6.73 | -- | |||
Key definitions
Frequently Asked Questions
Common questions about Nebius
Nebius offers 30 models that we track: Kimi K3 (max), GLM-5.2 (max) (FP4), MiniMax-M3 (FP8), Kimi K2.6, Kimi K2.7 Code (FP4), GLM-5.1 (FP8, Base), Nemotron 3 Ultra, GLM-5.1 (FP8, Base), Kimi K2.6, GLM-5.2 (FP4), Qwen3.5 397B A17B (Base, FP4), MiniMax-M2.5 (FP4), Qwen3.5 397B A17B (Base, FP4), NVIDIA Nemotron 3 Super, gpt-oss-120b (high) (Base), Qwen3 235B 2507, Qwen3 Next 80B A3B (FP8), Nemotron 3 Nano Omni 30B A3B Reasoning (FP8), gpt-oss-120b (low) Base, NVIDIA Nemotron 3 Nano, Qwen3 32B Base, Hermes 4 70B (FP8), Llama 3.3 70B Base, Llama Nemotron Ultra Base, Qwen3 30B A3B 2507, Hermes 4 405B (FP8), Hermes 4 405B (FP8), Qwen3 32B Base, Gemma 3 27B (FP8), and Hermes 4 70B (FP8).
The most intelligent model available on Nebius is Kimi K3 (max) with an Intelligence Index score of 57.
The fastest model on Nebius by output speed is GLM-5.2 (max) (FP4) at 352.9 tokens per second.
The model with the lowest time to first answer token on Nebius is Qwen3 30B A3B 2507 at 1.12s. Lower latency means faster initial response time.
The most affordable model on Nebius by blended price is Nemotron 3 Nano Omni 30B A3B Reasoning (FP8) at $0.08 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on Nebius vary up to 54x across models, from $0.08 per 1M tokens for Nemotron 3 Nano Omni 30B A3B Reasoning (FP8) to $4.20 per 1M tokens for Kimi K3 (max).
Yes, Nebius offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
Yes, all 30 models on Nebius support JSON mode for structured output.
Yes, all 30 models on Nebius support function calling (tool use).
Yes, Nebius offers 19 reasoning models: Kimi K3 (max), GLM-5.2 (max) (FP4), MiniMax-M3 (FP8), Kimi K2.6, Kimi K2.7 Code (FP4), GLM-5.1 (FP8, Base), Nemotron 3 Ultra, Qwen3.5 397B A17B (Base, FP4), MiniMax-M2.5 (FP4), NVIDIA Nemotron 3 Super, gpt-oss-120b (high) (Base), Qwen3 Next 80B A3B (FP8), Nemotron 3 Nano Omni 30B A3B Reasoning (FP8), gpt-oss-120b (low) Base, NVIDIA Nemotron 3 Nano, Qwen3 32B Base, Hermes 4 70B (FP8), Llama Nemotron Ultra Base, and Hermes 4 405B (FP8). Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, all 30 models on Nebius are open weight models.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
When choosing a model on Nebius, consider: intelligence (for quality-sensitive tasks), output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and features like context window size, JSON mode, or function calling support.