Alibaba Cloud: Models Intelligence, Performance & Price
This analysis is intended to support you in choosing the best model provided by Alibaba Cloud for your use-case.
Most Intelligent
Intelligence Index
Total 28 models
Lowest Time per Task
Time per Intelligence Index task (minutes)
Total 28 models
Lowest Price
Blended price (per 1M tokens)
Total 28 models
Alibaba Cloud offers 28 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 Alibaba Cloud are Qwen4 Max (43), Qwen3.8 Max (0902) (34), and Qwen3.8-Flash-Next (29).
- For time per task, Qwen3.8 27B (non-reasoning) (18.5 min), Qwen3.8 27B (low) (28.7 min), and Qwen3.8-Flash-Next (29.3 min) finish an Intelligence Index task fastest on Alibaba Cloud.
- For latency, Qwen3 Omni 30B A3B (1.14s), Qwen3.5 397B A17B (non-reasoning) (2.40s), and Qwen3.5 Omni Plus (2.69s) offer the lowest time to first answer token.
- For pricing, Qwen3.8-Flash-Next ($0.09), Qwen3.5 Omni Flash ($0.17), and Qwen3.7 Plus ($0.30) offer the lowest blended prices per 1M tokens. Prices vary up to 3.6x across models.
- For context window size, Qwen3.8-Flash-Next (1M), Qwen3.8 27B (xhigh) (1M), and Qwen3.8 27B (medium) (1M) support the largest context windows on Alibaba Cloud.
Intelligence
Artificial Analysis Intelligence Index
Evaluations
Agentic knowledge work, Elo
Knowledge
1 - hallucination rate
Agentic SaaS workflows
Agentic scientific research workflows in a terminal
Agentic software engineering in open-source repositories
Reasoning & knowledge
Professional document reasoning, All-pass
Agentic coding & terminal use
Physics reasoning
Quantitative analysis on spreadsheets & documents
Legal agentic work, Hallucination-Gated All-Pass Rate
Agentic business operations
Agentic real-world work tasks, Elo
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
Pricing
Cost per Intelligence Index Task
Intelligence Index vs. Price
Speed
Average of the single-request and multi-turn agentic workloads measured
Time per Intelligence Index Task
Quality
Context Window
Latency
Average of the single-request and multi-turn agentic workloads measured
Time to First Answer Token
Reliability
Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured
Endpoint Error Rate
Caching
Cache Hit Rate
Cost per Task vs. Cache Hit Rate
Further Analysis | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Qwen3.8 Max (0902) | 34 | $10.02 | 48.6 | 114.90 | 0.0% | 984k | Proprietary | |||
Qwen3.8-Flash-Next | 29 | $0.57 | 33.8 | 92.56 | 0.0% | 1M | Open | |||
Qwen3.8 27B (xhigh) | 20 | $3.53 | 42.7 | 136.50 | 0.9% | 1M | Open | |||
Qwen3.8 27B (medium) | 17 | $2.16 | 32.8 | 29.89 | 0.0% | 1M | Open | |||
Qwen3.8 27B (low) | 16 | $1.91 | 30.4 | 21.65 | 0.0% | 1M | Open | |||
Qwen3.8 27B (non-reasoning) | 10 | $3.77 | 18.3 | 5.12 | 0.0% | 1M | Open | |||
Qwen3.5 Omni Plus | -- | -- | -- | 2.79 | 0.0% | 256k | Proprietary | |||
Qwen3.5 122B A10B (non-reasoning) | -- | -- | -- | 4.41 | 0.0% | 262k | Open | |||
Qwen3 Omni 30B A3B | -- | -- | -- | 1.13 | 0.0% | 66k | Open | |||
Qwen3.5 397B A17B | -- | -- | -- | 41.48 | 0.9% | 262k | Open | |||
Qwen3.5 27B | -- | -- | -- | 37.15 | 2.6% | 262k | Open | |||
Qwen3.5 122B A10B | -- | -- | -- | 18.49 | 0.0% | 262k | Open | |||
Qwen3.5 397B A17B (non-reasoning) | -- | -- | -- | 2.44 | 0.0% | 262k | Open | |||
Qwen3.6 27B (non-reasoning) | -- | -- | -- | 7.20 | 0.0% | 262k | Open | |||
Qwen3.5 35B A3B | -- | -- | -- | 16.09 | 2.2% | 262k | Open | |||
Qwen3.5 27B (non-reasoning) | -- | -- | -- | 11.73 | 0.0% | 262k | Open | |||
Qwen3.8 Max | -- | -- | -- | 127.62 | 0.0% | 1M | Proprietary | |||
Qwen3.6 35B A3B (non-reasoning) | -- | -- | -- | 3.15 | 0.0% | 262k | Open | |||
Qwen3.7 Max | -- | -- | -- | 13.84 | 0.0% | 991k | Proprietary | |||
Qwen3.5 Omni Flash | -- | -- | -- | 2.84 | 0.7% | 256k | Proprietary | |||
Qwen3.6 35B A3B | -- | -- | -- | 19.42 | 0.0% | 229k | Open | |||
Qwen3.6 27B | -- | -- | -- | 45.41 | 0.0% | 262k | Open | |||
Qwen3.8 2.4T A95B | -- | -- | -- | 119.52 | 1.8% | 262k | Open | |||
Qwen3.7 Plus | -- | -- | -- | 48.76 | 0.0% | 1M | Proprietary | |||
Qwen3 Omni 30B A3B | -- | -- | -- | 25.50 | 0.0% | 66k | Open | |||
Qwen3.5 35B A3B (non-reasoning) | -- | -- | -- | 3.63 | 0.0% | 262k | Open | |||
Qwen3.6 Plus | -- | -- | -- | 44.86 | 0.0% | 1M | Proprietary | |||
Key definitions
Frequently Asked Questions
Common questions about Alibaba Cloud
Alibaba Cloud offers 28 models that we track: Qwen4 Max, Qwen3.8 Max (0902), Qwen3.8-Flash-Next, Qwen3.8 27B (xhigh), Qwen3.8 27B (medium), Qwen3.8 27B (low), Qwen3.8 27B (non-reasoning), Qwen3.5 Omni Plus, Qwen3.5 122B A10B (non-reasoning), Qwen3 Omni 30B A3B, Qwen3.5 397B A17B, Qwen3.5 27B, Qwen3.5 122B A10B, Qwen3.5 397B A17B (non-reasoning), Qwen3.6 27B (non-reasoning), Qwen3.5 35B A3B, Qwen3.5 27B (non-reasoning), Qwen3.8 Max, Qwen3.6 35B A3B (non-reasoning), Qwen3.7 Max, Qwen3.5 Omni Flash, Qwen3.6 35B A3B, Qwen3.6 27B, Qwen3.8 2.4T A95B, Qwen3.7 Plus, Qwen3 Omni 30B A3B, Qwen3.5 35B A3B (non-reasoning), and Qwen3.6 Plus.
The most intelligent model available on Alibaba Cloud is Qwen4 Max with an Intelligence Index score of 43.
The model with the lowest time per task on Alibaba Cloud is Qwen3.8 27B (non-reasoning) at 18.5 minutes per Intelligence Index task. Time per task is the output tokens per task divided by end-to-end output speed, so lower means a task finishes sooner.
The model with the lowest time to first answer token on Alibaba Cloud is Qwen3 Omni 30B A3B at 1.14s. Lower latency means faster initial response time.
The most affordable model on Alibaba Cloud by blended price is Qwen3.8-Flash-Next at $0.0882 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on Alibaba Cloud vary up to 18.1x across models, from $0.0882 per 1M tokens for Qwen3.8-Flash-Next to $1.60 per 1M tokens for Qwen3.7 Max.
Yes, Alibaba Cloud offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
Yes, all 28 models on Alibaba Cloud support JSON mode for structured output.
Yes, all 28 models on Alibaba Cloud support function calling (tool use).
Yes, Alibaba Cloud offers 18 reasoning models: Qwen4 Max, Qwen3.8 Max (0902), Qwen3.8-Flash-Next, Qwen3.8 27B (xhigh), Qwen3.8 27B (medium), Qwen3.8 27B (low), Qwen3.5 397B A17B, Qwen3.5 27B, Qwen3.5 122B A10B, Qwen3.5 35B A3B, Qwen3.8 Max, Qwen3.7 Max, Qwen3.6 35B A3B, Qwen3.6 27B, Qwen3.8 2.4T A95B, Qwen3.7 Plus, Qwen3 Omni 30B A3B, and Qwen3.6 Plus. Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, 20 of 28 models on Alibaba Cloud are open weight models: Qwen3.8-Flash-Next, Qwen3.8 27B (xhigh), Qwen3.8 27B (medium), Qwen3.8 27B (low), Qwen3.8 27B (non-reasoning), Qwen3.5 122B A10B (non-reasoning), Qwen3 Omni 30B A3B, Qwen3.5 397B A17B, Qwen3.5 27B, Qwen3.5 122B A10B, Qwen3.5 397B A17B (non-reasoning), Qwen3.6 27B (non-reasoning), Qwen3.5 35B A3B, Qwen3.5 27B (non-reasoning), Qwen3.6 35B A3B (non-reasoning), Qwen3.6 35B A3B, Qwen3.6 27B, Qwen3.8 2.4T A95B, Qwen3 Omni 30B A3B, and Qwen3.5 35B A3B (non-reasoning).
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 Alibaba Cloud, consider: intelligence (for quality-sensitive tasks), time per task (how quickly the model completes an Intelligence Index task), 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.