Qwen3-Coder Models by Alibaba
10 models2025–2026Up to 1M ctxFrom $0.1/1M input
About
Qwen3-Coder is a family of 10 AI models by Alibaba, released between 2025 and 2026.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
10 in view
Qwen3-Coder-NextCurrent
Use when the workload needs code, 256K context, and reasoning.
2026-02code256K contextreasoning
Qwen3-Coder-30B-A3B-InstructCurrent
Use when the workload needs code, structured outputs, and code execution.
2025-12codestructured outputscode execution
Use when the workload needs code, 256K context, and 480B parameters.
2025-01code256K context480B parameters
Qwen3-Coder-8B-InstructCurrent
Use when the workload needs code, 256K context, and 8B parameters.
2025-01code256K context8B parameters
Qwen3-Coder-14B-InstructCurrent
Use when the workload needs code, 256K context, and 14B parameters.
2025-01code256K context14B parameters
Qwen3-Coder-32B-InstructCurrent
Use when the workload needs code, 256K context, and 32B parameters.
2025-01code256K context32B parameters
Qwen3-Coder-3B-InstructCurrent
Use when the workload needs code, 256K context, and 3B parameters.
2025-01code256K context3B parameters
Qwen3-Coder-72B-InstructCurrent
Use when the workload needs code, 256K context, and 72B parameters.
2025-01code256K context72B parameters
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Qwen3-Coder-Next | Use when the workload needs code, 256K context, and reasoning. | 2026-02 | code256K contextreasoning | Current |
| Qwen3-Coder-30B-A3B-Instruct | Use when the workload needs code, structured outputs, and code execution. | 2025-12 | codestructured outputscode execution | Current |
| Qwen3-Coder-Plus | Use when the workload needs 1M context. | 2025-09 | 1M context | Current |
| Qwen3-Coder-Flash | Use when the workload needs 1M context. | 2025-07 | 1M context | Current |
| Qwen3-Coder-480B-A35B-Instruct | Use when the workload needs code, 256K context, and 480B parameters. | 2025-01 | code256K context480B parameters | Current |
| Qwen3-Coder-8B-Instruct | Use when the workload needs code, 256K context, and 8B parameters. | 2025-01 | code256K context8B parameters | Current |
| Qwen3-Coder-14B-Instruct | Use when the workload needs code, 256K context, and 14B parameters. | 2025-01 | code256K context14B parameters | Current |
| Qwen3-Coder-32B-Instruct | Use when the workload needs code, 256K context, and 32B parameters. | 2025-01 | code256K context32B parameters | Current |
| Qwen3-Coder-3B-Instruct | Use when the workload needs code, 256K context, and 3B parameters. | 2025-01 | code256K context3B parameters | Current |
| Qwen3-Coder-72B-Instruct | Use when the workload needs code, 256K context, and 72B parameters. | 2025-01 | code256K context72B parameters | Current |
Release Timeline
5 release groups2026-02
1 current
Qwen3-Coder-Next
Currentcode256K contextreasoning
2025-12
1 current
Qwen3-Coder-30B-A3B-Instruct
Currentcodestructured outputscode execution
2025-09
1 current
Qwen3-Coder-Plus
Current1M context
2025-07
1 current
Qwen3-Coder-Flash
Current1M context
2025-01
6 current
Qwen3-Coder-14B-Instruct
Currentcode256K context14B parameters
Qwen3-Coder-32B-Instruct
Currentcode256K context32B parameters
Qwen3-Coder-3B-Instruct
Currentcode256K context3B parameters
Qwen3-Coder-480B-A35B-Instruct
Currentcode256K context480B parameters
Qwen3-Coder-72B-Instruct
Currentcode256K context72B parameters
Qwen3-Coder-8B-Instruct
Currentcode256K context8B parameters
Specifications(10 models)
| Model | Released | Context | Parameters | Reasoning | Fn Calling | Tool Use | Structured Outputs | Code Exec |
|---|---|---|---|---|---|---|---|---|
| Qwen3-Coder-Next | 2026-02 | 256K | 80B total, 3B active | Yes | Yes | Yes | Yes | Yes |
| Qwen3-Coder-30B-A3B-Instruct | 2025-12 | — | — | No | No | No | Yes | Yes |
| Qwen3-Coder-Plus | 2025-09 | 1M | — | No | No | No | No | No |
| Qwen3-Coder-Flash | 2025-07 | 1M | — | No | No | No | No | No |
| Qwen3-Coder-480B-A35B-Instruct | 2025-01 | 256K | 480B | No | No | No | Yes | Yes |
| Qwen3-Coder-8B-Instruct | 2025-01 | 256K | 8B | No | No | No | No | Yes |
| Qwen3-Coder-14B-Instruct | 2025-01 | 256K | 14B | No | No | No | No | Yes |
| Qwen3-Coder-32B-Instruct | 2025-01 | 256K | 32B | No | No | No | No | Yes |
| Qwen3-Coder-3B-Instruct | 2025-01 | 256K | 3B | No | No | No | No | Yes |
| Qwen3-Coder-72B-Instruct | 2025-01 | 256K | 72B | No | No | No | No | Yes |
Available From(6 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Qwen3-Coder-3B-Instruct | Fireworks AI | $0.1 | $0.1 | Serverless |
| Qwen3-Coder-Next | OpenRouter | $0.12 | $0.8 | Serverless |
| Qwen3-Coder-30B-A3B-Instruct | AWS Bedrock | $0.15 | $0.62 | Serverless |
| Qwen3-Coder-8B-Instruct | Fireworks AI | $0.2 | $0.2 | Serverless |
| Qwen3-Coder-14B-Instruct | Fireworks AI | $0.2 | $0.2 | Serverless |
| Qwen3-Coder-480B-A35B-Instruct | GCP Vertex AI | $0.22 | $1.8 | Serverless |
| Qwen3-Coder-Next | AWS Bedrock | $0.5 | $1.2 | Serverless |
| Qwen3-Coder-32B-Instruct | Fireworks AI | $0.9 | $0.9 | Serverless |
| Qwen3-Coder-72B-Instruct | Fireworks AI | $0.9 | $0.9 | Serverless |
| Qwen3-Coder-480B-A35B-Instruct | Fireworks AI | $1.2 | $1.2 | Serverless |
| Qwen3-Coder-Flash | Alibaba Cloud PAI-EAS | $1.6 | $9.6 | Serverless |
Frequently Asked Questions
- What is Qwen3-Coder used for?
- Qwen3-Coder is used for code, reasoning, and agent workflows and tool use. The family description and listed model capabilities point to those workloads as the best fit.
- How does Qwen3-Coder compare to Tongyi DeepResearch?
- Qwen3-Coder by Alibaba is strongest where you need code, while Tongyi DeepResearch by Alibaba is the closest related family to check for adjacent model selection. Qwen3-Coder has 10 listed variants and reaches up to 1M context, while Tongyi DeepResearch reaches up to 131K context, so compare the specs and pricing tables before choosing a production model.
- Which Qwen3-Coder model should I use?
- For the lowest listed input price, start with Qwen3-Coder-3B-Instruct through Fireworks AI at $0.1/1M input tokens. For the most capable/latest local choice, evaluate Qwen3-Coder-Next with 256K context and reasoning, tool use, function calling, and structured outputs.
Models(10)
Qwen3-Coder-Next
2026-02256K80B total, 3B active2 providers
Reasoning
Qwen3-Coder-30B-A3B-Instruct
2025-121 provider
Qwen3-Coder-Plus
2025-091M1 provider
Qwen3-Coder-Flash
2025-071M1 provider
Qwen3-Coder-480B-A35B-Instruct
2025-01256K480B4 providers
Qwen3-Coder-8B-Instruct
2025-01256K8B1 provider
Qwen3-Coder-14B-Instruct
2025-01256K14B1 provider
Qwen3-Coder-32B-Instruct
2025-01256K32B1 provider
Qwen3-Coder-3B-Instruct
2025-01256K3B1 provider
Qwen3-Coder-72B-Instruct
2025-01256K72B1 provider