Alpha
Contact
Use · DeepSeek-R1

Is it good enough?

Ownership levelSubstantialnone·limited·partial·substantial·fullAnalytical input C ยท 64.4/100

This page is a projection of the one entry record, the Reliability factor that Use covers. The full verdict is set by all four factors together, floor-weighted so the weakest caps the whole.

Which domain expands which factor
  • AssessUse & modify + Transparency
  • ImplementData control + Reliability
  • UseReliability
  • SupportTransparency

Capabilities & modalities

R1 is a text-only reasoning model with strong maths, coding and analysis. It emits explicit reasoning traces (think tags) before its final answer, which distinguishes it from the DeepSeek-V3 general-instruct line.

Context window & long-context behaviour

The model card declares a 128K-token context. Effective long-context recall is unmeasured, so treat 128K as the declared window and working depth as unverified.

Prompt format & chat template

R1 ships a chat template with a reasoning / think-tag convention. Use apply_chat_template instead of hand-assembling prompts. Handle the reasoning segment at the client and decide whether to expose or strip it before showing users.

Language coverage

Strong English and Chinese. Per-language depth varies and is not exhaustively documented, so evaluate for the target languages.

Function / tool calling

Native tool calling is not documented as a primary behaviour of R1 and is unverified. Confirm support per serving stack and test it before relying on it. Do not assume parity with the V3 instruct line.

Structured / JSON-constrained output

The model documents no native JSON or schema-constrained output guarantee. Constrained decoding is available at the serving layer (vLLM and SGLang grammars) as a runtime feature of the server.

How this scores

The ownership factor this domain covers, drawn from the one entry record.

3

ReliabilityIs it reliable and good enough for the job?

Strong

Performance 4, operational 5 and safety 3: a strong reasoner with first-class serving. All three score at least 3 and two score at least 4, so reliability is strong. No first-party guard model ships.

How this scores (AOI sub-dimensions)
Operational5/5how practical it is to run, serve and maintain in productionServing is supported across vLLM, SGLang, llama.cpp and Ollama.
Safety3/5whether misuse risks are evaluated and guardrails are providedReleased as reasoning variants with documented behaviour, which meets the score-3 anchor.
What this means for adoptionSelf-hosted DeepSeek-R1 is substantially owned. MIT permits commercial use, modification and distillation, and self-hosting keeps data with the deployer, so use-and-modify and data-control are both strong. Ownership stops short of full because the training corpus and code are closed, the weights carry China-aligned filtering that cannot be inspected, and no first-party guard model exists. Deploy behind your own guardrails, and treat EU high-stakes use as needing a self-assembled compliance package.

Sources

The same evidence records as the entry sheet. Read means the text was verified; unverified means it is known to exist but not yet read.

Licenceread2026-11-19
The DeepSeek-R1 README states: "This code repository and the model weights are licensed under the MIT License.
Model cardread2026-11-19
The model card on the verified deepseek-ai Hugging Face org lists a 671B-total / 37B-active MoE, 128K context and safetensors.
Third-party analysisunverified2026-08-03
On public third-party leaderboards DeepSeek-R1 is among the strongest open-weight reasoning models (maths, coding, analysis).
Third-party analysisunverified2026-08-03
Independent analysis notes DeepSeek open-weight models apply China-aligned content filtering on politically sensitive topics, with lighter safety tuning than Western frontier labs and no companion guard model.
Third-party analysisunverified2026-08-03
No public EU AI Act training-content summary, copyright policy, or GPAI documentation package is published for DeepSeek-R1, and the training corpus is not released.