DeepSeek V3.2

Tool callingStructured outputReasoning

Hugging Face Launch post Website

Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use

Context: 164K
Released: 10mo ago
Parameters: —
Input / 1M tokens: $0.26
Output / 1M tokens: $0.80
Cached input / 1M tokens: $0.14

Sources:models.devOpenRouter

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

Other published benchmarks

Evaluation	Reported score and source
ARC-AGI-2	4.0% Third-party evaluator · deepseek/deepseek-v3.2 External evaluation via Epoch AI Original-source terms apply
Terminal Bench	39.6% https://www.tbench.ai/leaderboard/terminal-bench/2.0 · deepseek/deepseek-v3.2 Agent: Terminus 2 https://www.tbench.ai/leaderboard/terminal-bench/2.0 Original-source terms apply

Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.

Specifications

Developer: DeepSeek
Model family: DeepSeek
Type: Language model
Predecessor: —
Successor: —
Released: December 1, 2025
Model ID: deepseek/deepseek-v3.2
Openness: Open weights
License: MIT License
Parameters: —
Input modalities: text
Output modalities: text
Modality: text->text
Tokenizer: DeepSeek
Max output tokens: 147K
Knowledge cutoff: 2024-07
Hugging Face: deepseek-ai/DeepSeek-V3.2
Source freshness: Metadata synced Oct 8, 2026, 10:17 PM UTC

Additional source evidence

Source model: deepseek/deepseek-v3.2
Source family: deepseek
Last updated: 2025-12-01
Knowledge cutoff: 2024-07
Model context: 128K
Model output limit: 64K
Open weights: Yes
File attachments: No
Reasoning: Yes
Tool calling: Yes
Structured output: Yes
Temperature control: Yes

Downloadable weights

Hugging Face

Complete source record (JSON)

{
  "id": "deepseek/deepseek-v3.2",
  "name": "DeepSeek V3.2",
  "description": "Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use",
  "family": "deepseek",
  "last_updated": "2025-12-01",
  "knowledge": "2024-07",
  "attachment": false,
  "reasoning": true,
  "tool_call": true,
  "structured_output": true,
  "temperature": true,
  "modalities": {
    "input": [
      "text"
    ],
    "output": [
      "text"
    ]
  },
  "limit": {
    "context": 128000,
    "output": 64000
  },
  "release_date": "2025-12-01",
  "open_weights": true,
  "license": "MIT License",
  "weights": [
    {
      "label": "Hugging Face",
      "url": "https://huggingface.co/deepseek-ai/DeepSeek-V3.2"
    }
  ]
}

Infrastructure

Provider availability

All providers

Provider	Endpoint	Context	Input /M tokens	Output /M tokens	Cached input /M tokens	Cache write /M tokens	Uptime 1d	Data handling
GMICloud	gmicloud/fp8	164K	$0.21	$0.31	$0.02	—	100.00%	ZDR: No Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
SiliconFlow	siliconflow/fp8	164K	$0.26	$0.80	$0.14	—	99.51%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
AtlasCloud	atlas-cloud/fp8	164K	$0.26	$0.38	$0.13	—	90.35%	ZDR: No Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
DeepInfra	deepinfra/fp4	164K	$0.26	$0.38	$0.13	—	95.83%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
Venice	venice	160K	$0.27	$0.39	$0.13	—	94.75%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
Baidu	baidu/fp8	131K	$0.28	$0.42	$0.03	—	99.93%	ZDR: No Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
DigitalOcean	digitalocean	164K	$0.30	$0.96	$0.09	—	99.93%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
Friendli	friendli	164K	$0.50	$1.50	$0.25	—	99.93%	ZDR: No Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
Google	google-vertex	164K	$0.56	$1.68	—	—	98.23%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
Phala	phala	164K	$1.00	$1.00	$0.50	—	99.73%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
Mara	mara	33K	$3.00	$4.50	—	—	86.14%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15
SambaNova	sambanova	33K	$3.00	$4.50	—	—	90.00%	ZDR: Yes Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15

Provider data:OpenRouter

Deployment graph

Can I run this?

Vertical Tabs only shows compatibility when it has structured evidence tying this exact model to hardware, runtime, quantization, memory use, and observed performance.

Browse hardware