Codesota · Models · DeepSeek-OCR (Small, 100 vision tokens)DeepSeek1 results · 1 benchmarks
Model card

DeepSeek-OCR (Small, 100 vision tokens).

DeepSeekopen-sourceOCR / document VLMcheck model card1 current SOTA

Added from Papers With Code public API dump for Fox (English subset, 600-1300 text tokens).

§ 02 · Benchmarks

Every benchmark DeepSeek-OCR (Small, 100 vision tokens) has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01Fox English OCR subset—precision98.5%#1/12025-10-21source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 05 · Related models

Other DeepSeek models scored on Codesota.

DeepSeek R1
671B MoE params · 19 results
DeepSeek-V3
15 results
DeepSeek-V3.2
14 results
DeepSeek-V3.2-Speciale
11 results
DeepSeek-R1-0528
10 results
DeepSeek-Coder-V2-Instruct
Unknown params · 9 results
DeepSeek-v3-0324
9 results
DeepSeek-v3.1 (no thinking)
7 results
§ 06 · Sources & freshness

Where these numbers come from.

paperswithcode-public-api
1
result
1 of 1 rows marked verified.