Scaleway: Models Intelligence, Performance & Price
This analysis is intended to support you in choosing the best model provided by Scaleway for your use-case.
Most Intelligent
GLM-5.3 (max)
DeepSeek V4.1 Flash (max)GLM-5.3 (low)GLM-5.2 (max)DeepSeek V4.1 Flash (non-reasoning)Intelligence Index
Total 13 models
Fastest
DeepSeek V4.1 Flash (max)DeepSeek V4.1 Flash (non-reasoning)GLM-5.3 (low)
Qwen3 Coder 30B A3BGLM-5.3 (max)Output speed
Total 13 models
Lowest Price
gpt-oss-120b (high)
Gemma 4 26B A4B (non-reasoning)Qwen3 Coder 30B A3BQwen3.6 35B A3BQwen3.6 35B A3B (non-reasoning)Blended price (per 1M tokens)
Total 13 models
Scaleway offers 13 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 Scaleway are GLM-5.3 (max) (45), DeepSeek V4.1 Flash (max) (39), and GLM-5.3 (low) (34).
- For output speed, the fastest models are DeepSeek V4.1 Flash (max) (254 t/s), DeepSeek V4.1 Flash (non-reasoning) (223 t/s), and GLM-5.3 (low) (220 t/s).
- For latency, Gemma 4 26B A4B (non-reasoning) (0.85s), Qwen3 Coder 30B A3B (1.02s), and GLM-5.2 (non-reasoning) (1.17s) offer the lowest time to first answer token.
- For pricing, gpt-oss-120b (high) ($0.22), Gemma 4 26B A4B (non-reasoning) ($0.28), and Qwen3 Coder 30B A3B ($0.30) offer the lowest blended prices per 1M tokens.
- For context window size, GLM-5.3 (max) (262k), DeepSeek V4.1 Flash (max) (262k), and GLM-5.3 (low) (262k) support the largest context windows on Scaleway.
Intelligence Evaluations
Artificial Analysis Intelligence Index
Evaluations
Agentic knowledge work, Elo
Knowledge
1 - hallucination rate
Agentic SaaS workflows
Agentic real-world work tasks, Elo
Reasoning & knowledge
Professional document reasoning, All-pass
Agentic coding & terminal use
Long context reasoning
Coding
Physics reasoning
Agentic scientific research workflows in a terminal
Agentic software engineering in open-source repositories
Quantitative analysis on spreadsheets & documents
Legal agentic work, Hallucination-Gated All-Pass Rate
Agentic business operations
Agentic interactive 3D experiences, Elo
Product and brand videos designed in code, Elo
Kubernetes incident root-cause analysis
Medical long context reasoning
Visual reasoning
Rich, varied, well-crafted language
Well-structured, scannable, answer-first
Truthful, acknowledges uncertainty, non-sycophantic
Use of filler and generic language
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
Cache Behaviour
Cache Hit Rate
Cost per Task vs. Cache Hit Rate
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 | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
GLM-5.3 (max) | 262k | Open | 45 | $12.94 | 201 | 1.37 | 13.82 | 9.96 | |||
DeepSeek V4.1 Flash (max) | 262k | Open | 39 | $4.90 | 254 | 1.37 | 11.22 | 7.88 | |||
GLM-5.3 (low) | 262k | Open | 34 | $5.46 | 220 | 1.30 | 12.69 | 9.11 | |||
GLM-5.2 (max) | 262k | Open | 34 | $4.37 | 165 | 1.13 | 16.27 | 12.11 | |||
DeepSeek V4.1 Flash (non-reasoning) | 262k | Open | 25 | $3.42 | 223 | 1.34 | 3.57 | -- | |||
GLM-5.2 (non-reasoning) | 262k | Open | 22* | -- | 159 | 1.17 | 4.32 | -- | |||
Qwen3.6 35B A3B | 262k | Open | 18 | $0.37 | 153 | 1.14 | 39.69 | 35.27 | |||
Qwen3.6 35B A3B (non-reasoning) | 262k | Open | 15* | -- | 138 | 1.18 | 4.80 | -- | |||
Gemma 4 26B A4B (non-reasoning) | 262k | Open | 13* | -- | 177 | 0.85 | 3.67 | -- | |||
Qwen3 235B 2507 | 250k | Open | 12* | -- | 56 | 1.51 | 10.42 | -- | |||
gpt-oss-120b (high) | 128k | Open | 12 | $0.12 | 140 | 1.33 | 19.20 | 14.30 | |||
Qwen3 Coder 30B A3B | 128k | Open | 10* | -- | 215 | 1.02 | 3.35 | -- | |||
 Llama 3.3 70B | 100k | Open | 8* | -- | 85 | 1.38 | 7.24 | -- | |||
Key definitions
Frequently Asked Questions
Common questions about Scaleway
Scaleway offers 13 models that we track: GLM-5.3 (max), DeepSeek V4.1 Flash (max), GLM-5.3 (low), GLM-5.2 (max), DeepSeek V4.1 Flash (non-reasoning), GLM-5.2 (non-reasoning), Qwen3.6 35B A3B, Qwen3.6 35B A3B (non-reasoning), Gemma 4 26B A4B (non-reasoning), Qwen3 235B 2507, gpt-oss-120b (high), Qwen3 Coder 30B A3B, and Llama 3.3 70B.
The most intelligent model available on Scaleway is GLM-5.3 (max) with an Intelligence Index score of 45.
The fastest model on Scaleway by output speed is DeepSeek V4.1 Flash (max) at 253.8 tokens per second.
The model with the lowest time to first answer token on Scaleway is Gemma 4 26B A4B (non-reasoning) at 0.85s. Lower latency means faster initial response time.
The most affordable model on Scaleway by blended price is gpt-oss-120b (high) at $0.22 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on Scaleway vary up to 11x across models, from $0.22 per 1M tokens for gpt-oss-120b (high) to $2.52 per 1M tokens for GLM-5.2 (non-reasoning).
Yes, Scaleway offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
Yes, all 13 models on Scaleway support JSON mode for structured output.
Yes, all 13 models on Scaleway support function calling (tool use).
Yes, Scaleway offers 6 reasoning models: GLM-5.3 (max), DeepSeek V4.1 Flash (max), GLM-5.3 (low), GLM-5.2 (max), Qwen3.6 35B A3B, and gpt-oss-120b (high). Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, all 13 models on Scaleway 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 Scaleway, 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.