deepseek-ai/DeepSeek-V3
DeepInfra Inc./🇺🇸 US/chat
DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures.MoreLess
Which id to call
deepinfra/deepseek-ai/DeepSeek-V3This exact deployment on DeepInfra Inc., with no routing and no failover. Send it as the model field.
Input /1M
$0.85
DeepInfra Inc.
Output /1M
$0.90
1.1x input
Context
128K
8K output
Added
Dec 2024
chat
Capabilities 2/8
Provider rates
What deepseek-ai/DeepSeek-V3 costs
Provider prices per 1M tokens, updated September 11, 2026.moreless
These are the upstream provider rates. Pay as you go adds 5%, or 0% if you bring your own keys, and there is no per-request fee. Prompt caching and routing change what you pay against these rates, not the rates themselves.
Input /1M
$0.85
Output /1M
$0.90
Cache write /1M
-
Cache read /1M
-
What a workload costs
- 100K input + 10K output
- $0.0940
- 1M input + 100K output
- $0.94
- 10M input + 1M output
- $9.40
At the rates above. This endpoint does not offer prompt caching, so repeated context is billed as new input every time.
OpenAI compatible
Call it in three lines
Change the base url, use your Requesty key, set the model to deepinfra/deepseek-ai/DeepSeek-V3. Existing OpenAI SDK code needs no other edit.
123456789101112131415from openai import OpenAI client = OpenAI( api_key="YOUR_REQUESTY_API_KEY", base_url="https://router.requesty.ai/v1", ) response = client.chat.completions.create( model="deepinfra/deepseek-ai/DeepSeek-V3", messages=[ {"role": "user", "content": "Explain quantum computing in one paragraph."}, ], ) print(response.choices[0].message.content)
Reference
Specs and data terms
What the catalog reports for this deployment, and what the provider does with the traffic.
Released 2024-12-26
Benchmark scores
Benchmarks measure the model, not this endpoint, so they are the same wherever these weights are served.
Artificial Analysis Coding Index: a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.
Graduate-level physics, chemistry & biology questions designed to resist Googling.
Artificial Analysis Intelligence Index: a composite of multiple evaluations measuring overall model capability.
Scores from official model cards, Artificial Analysis and public leaderboards. They measure specific skills and do not capture every aspect of model quality, so test on your own workload.
Same provider
More from DeepInfra Inc.
Newest first, on the same provider and the same key.
Reference
deepseek-ai/DeepSeek-V3 questions
How much does deepseek-ai/DeepSeek-V3 cost?
deepseek-ai/DeepSeek-V3 is priced at $0.85 per million input tokens and $0.90 per million output tokens when accessed via Requesty. Those are the upstream provider rates: pay as you go adds 5% on top, or 0% if you bring your own provider keys.
What is the context window of deepseek-ai/DeepSeek-V3?
deepseek-ai/DeepSeek-V3 has a context window of 128K tokens, with a maximum output of 8K tokens per response. That's roughly 171 words of input you can fit in a single prompt.
How does deepseek-ai/DeepSeek-V3 perform on benchmarks?
deepseek-ai/DeepSeek-V3 scores 75.2% on MMLU Pro, 55.7% on GPQA Diamond, 35.9% on LiveCodeBench. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can deepseek-ai/DeepSeek-V3 do?
deepseek-ai/DeepSeek-V3 supports tool calling, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use deepseek-ai/DeepSeek-V3 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "deepinfra/deepseek-ai/DeepSeek-V3". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run deepseek-ai/DeepSeek-V3 through Requesty?
Yes. deepseek-ai/DeepSeek-V3 runs through Requesty's OpenAI-compatible API, served from DeepInfra Inc.. You do not host the model yourself: point base_url at Requesty, set the model to "deepinfra/deepseek-ai/DeepSeek-V3", and requests are routed to the upstream provider with automatic failover. The same key gives you 600+ other models too.
Call deepseek-ai/DeepSeek-V3 through one endpoint
One key for this endpoint and 600+ other models. No markup on provider prices, automatic failover, caching built in.