DeepSeek Chat
DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions.
deepseek-chatReleased 26 Dec 2024 (Dec 2024)Pricing
List prices at DeepSeek, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $0.28 | $0.42 | |
| Cached input 90% off input | $0.028 | 90% off input |
What it costs a month
At DeepSeek list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$92.82
- RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$89.04
- Coding agent1,000 requests a day, 40K in / 2K out, cached$150
- Overnight batch20,000 requests a day, 4K in / 0.4K out$773
Where to run it
2 hosts serve DeepSeek Chat. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
DeepSeekLab API deepseek-chat | $0.28 / $0.42 | deepseek-chat | |
Inference APIsGateway deepseek-chat | $0.19 / $0.38 −25% vs headline | deepseek-chat |
Show 1 more listings: regions, snapshots and variantsHide regional listings
| DeepSeek | |||||||
| Alias deepseek-chat | $0.28 / $0.42 | deepseek-chat | |||||
Capabilities
Core
- Streaming
- Function calling
- JSON mode
- Structured outputs
Multimodal
- Vision
- Image input
- Audio
- Image generation
- Video
Advanced
- Reasoning
- Tool calling
- MCP
- Prompt caching
- Batch API
- Embeddings
- Takes in
- Text
- Produces
- Text
- Features
- Parallel tool callsPrompt caching
- Open weights
- License: DeepSeek License (MIT code)Runs in about 1642.8 GB of GPU memoryRuntimes: SGLang, TGI, vLLM
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
- Inference APIs25% lessdeepseek-chat$0.19 / $0.38
Smaller DeepSeek models
No cheaper DeepSeek model keeps its key capabilities.
Other frontier labs
- 37% less
- 37% less
- 17% less
Open-weight picks
- 20% less
- 43% less
- 55% less
- 56% less
Benchmarks
Scores published on the model card.
- Aider Polyglot17.8%
Call it
Using DeepSeek's API.
from litellm import completion
response = completion(
model="deepseek/deepseek-chat",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does DeepSeek Chat cost?
On DeepSeek, DeepSeek Chat costs $0.28 per million input tokens and $0.42 per million output tokens. Cached input costs $0.028 per million tokens. The cheapest host is Inference APIs, at $0.19 per million input tokens and $0.38 per million output tokens.
Which APIs offer DeepSeek Chat?
DeepSeek Chat runs on 2 APIs: DeepSeek and Inference APIs.
What is DeepSeek Chat's context window?
131K tokens, with up to 8K tokens in one reply.
Is DeepSeek Chat being retired?
No retirement date is scheduled for it on any API we track.
What are cheaper alternatives to DeepSeek Chat?
Models that cost less and keep most of what it can do: Claude Haiku 5.5 on Anthropic (37% less), GPT-6 Luna on OpenAI (37% less) and Qwen3.8 27B on EmpirioLabs AI (20% less). DeepSeek Chat itself costs 25% less on Inference APIs. Savings are on a mix of 3 input tokens to 1 output token.
When was DeepSeek Chat released?
DeepSeek released it on 26 Dec 2024.
What is DeepSeek Chat's API model id?
Use deepseek-chat with DeepSeek.
Sources
Raw catalog entry for deepseek-chat on DeepSeek
{
"cache_read_input_token_cost": 2.8e-8,
"deprecation_date": "2026-07-24",
"deprecation_source": "https://api-docs.deepseek.com/updates",
"input_cost_per_token": 2.8e-7,
"litellm_provider": "deepseek",
"max_input_tokens": 131072,
"max_output_tokens": 8192,
"mode": "chat",
"output_cost_per_token": 4.2e-7,
"source": "https://api-docs.deepseek.com/quick_start/pricing",
"supported_endpoints": [
"/v1/chat/completions"
],
"supports_function_calling": true,
"supports_native_streaming": true,
"supports_parallel_function_calling": true,
"supports_prompt_caching": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_tool_choice": true
}