DeepSeek R1 0528 deepseek-r1-0528
deepseek-ai · deepseek reasoning flagship
unranked — no quality signals yet
- params
- 671B (A37B)
- arch
- moe
- context
- 160k
- license
- mit open weights
- released
- May 2025
- reasoning
- yes
- train compute
- 4.0e+24 FLOP
- downloads/30d
- 213.5k
# architecture
- layers
- 61
- d_model
- 7168
- heads
- 128q / 128kv
- head dim
- 56
- experts
- top-8 of 256 + 1 shared
- vocab
- 129280
- family
- deepseek_v3
signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.
# why ranked
Overall: provisional — score 51.0 ● low — a single signal; treat with caution
provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.
| signal | weight | input (0–100) |
|---|---|---|
| aa_intelligence | 0.70 | — |
| bench_composite | 0.30 | 51.0 |
benchmark panel evidence:
| complete panel | score (0–100) |
|---|---|
| aider-polyglot-v1 | 58.4 |
| aime-2025-v1 | 66.2 |
| browsecomp-v1 | 28.4 |
missing signals are dropped and the remaining weights renormalized — never imputed.
Coding: provisional — score 90.9 ● low — a single signal; treat with caution
provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.
| signal | weight | input (0–100) |
|---|---|---|
| aa_coding | 0.30 | — |
| aider_polyglot | 0.30 | 90.9 |
| swe_bench_verified | 0.40 | — |
missing signals are dropped and the remaining weights renormalized — never imputed.
# trend
methodology changed during this history window; score movement across that boundary is not model movement. See methodology v5.
# benchmarks
# where to run
| provider | quant | ctx | $/M in | $/M out | $/M cache | price src | tps | uptime | ✓ |
|---|---|---|---|---|---|---|---|---|---|
| Featherless | — | — | — | — | via hfrouter | — | — | ||
| Lambda | 128k | $0.20 | $0.60 | — | via litellm | — | — | ||
| DeepInfra | fp4 | 160k | $0.50 | $2.15 | — | deepinfra | — | 99.9% | ✓ |
| SiliconFlow | fp8 | 160k | $0.50 | $2.18 | — | via openrouter | — | 99.6% | |
| StreamLake | unknown | 125k | $0.57 | $2.29 | — | via openrouter | — | 99.9% | |
| Novita | fp8 | 160k | $0.70 | $2.50 | — | novita | — | 99.7% | ✓ |
| Nebius | 164k | $0.80 | $2.40 | — | via litellm | — | — | ||
| Hyperbolic | 160k | $3.00 | $3.00 | — | hyperbolic | — | — | ✓ | |
| Crusoe | 160k | $3.00 | $7.00 | — | via litellm | — | — |
sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price
source aliases
- aa
deepseek-r1-0528- aider
DeepSeek R1 (0528)- arena
DeepSeek-R1-0528- epoch
DeepSeek-R1 (May 2025)- openrouter
deepseek/deepseek-r1-0528