Together AI
Reasoning

DeepSeek-R1 (Together)

Reasoning

Open-weights reasoning model with chain-of-thought trained via RL. Comparable to o1 on math benchmarks at a fraction of the price.

DeepSeek-R1 (Together) is a reasoning AI model from Together AI. It costs $3.00 per million input tokens and $7.00 per million output tokens (blended $4.20/M), with a 64,000-token context window.

Profile inherited from upstream DeepSeek-R1 ↗ — this is a hosted variant of the same open-weights model.

RatesUSD per million tokens. Blend is 70% input, 30% output, our calculation.

INPUT
$3.00/M
per million input tokens
OUTPUT
$7.00/M
per million output tokens
BLENDED 70/30
$4.20/M
unchanged since 3 May
CONTEXT
64,000
tokens

Hand-maintained from Together AI's published pricing; no check date on record.

What it is good at
  • Open-weights reasoning
  • Visible chain-of-thought
  • Math & code
  • Cheap
Typical use cases
  • Reasoning agents
  • Math/STEM tools
  • Self-hosted reasoning

Benchmarks

vs. best public score
Scores inherited from DeepSeek-R1 — this is a hosted variant of the same open-weights model, so the underlying benchmark scores are identical.
MMLU90.8%
Multitask academic knowledge across 57 subjects. Source: DeepSeek (self-reported), MMLU (Pass@1), read 2026-11-10.
Graduate-level science questions, "Google-proof". Source: DeepSeek (self-reported), GPQA-Diamond (Pass@1), read 2026-11-10.
Real GitHub issues solved end-to-end. Source: DeepSeek (self-reported), SWE Verified (Resolved), read 2026-11-10.
LMArena Elo1398 Elo
Crowd-sourced head-to-head preference Elo rating. Source: LMArena (independent), LMArena Text Arena score, read 2026-11-10.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

Also reported by the provider

Results on benchmarks outside the set we compare across models. They are listed for completeness and are deliberately excluded from the bars above and from every ranking on this site, because a score is only comparable against the same benchmark variant. Why benchmark scores do not compare →

  • MATH-500 (pass@1)97.3%self-reported by DeepSeek · no date on the source · DeepSeek-R1 model card, pass@1 from 64 responses per query, temperature 0.6, top-p 0.95, up to 32,768 tokens. Not comparable with the MATH column, so shown here and never ranked.

More from Together AI

See all 10 →

Frequently asked questions

How much does DeepSeek-R1 (Together) cost?

DeepSeek-R1 (Together) costs $3.00 per million input tokens and $7.00 per million output tokens, for a blended reference rate of $4.20 per million tokens.

What is DeepSeek-R1 (Together)'s context window?

DeepSeek-R1 (Together) supports up to 64,000 tokens of context in a single request.

What is DeepSeek-R1 (Together) best for?

DeepSeek-R1 (Together) is well suited to Open-weights reasoning, Visible chain-of-thought and Math & code.

Who makes DeepSeek-R1 (Together)?

DeepSeek-R1 (Together) is developed and served by Together AI.

Terms used on this page