Best Local LLMs for OCR

Written by Jakub Rusinowski · Last updated August 15, 2026

Turning scanned pages, screenshots and photographed documents into accurate text.

Top pick: Gemma 4 27B ⭐

Scores 93.7/100 for OCR and document scanning. 27B parameters, needing about 17.1 GB at Q4_K_M, 125K context, Gemma License (commercial OK).

Ranked for OCR and document scanning

ModelScoreParamsContextLicenceQuality index
1. Gemma 4 27B ⭐93.727B125KGemma License (commercial OK)— (estimated)
2. Mistral Small 3.1 24B93.524B125KApache 2.0— (estimated)
3. Qwen 3.5 14B92.514B125KApache 2.0— (estimated)
4. Qwen 3.7 35B-A3B92.235B256KApache-2.0— (estimated)
5. Llama 3.2 11B Vision Instruct91.911B125KLlama Community— (estimated)
6. Qwen 2.5 VL 7B Instruct91.88B125KApache 2.0— (estimated)

Best pick for your memory budget

The strongest model overall is rarely the right answer — what matters is the strongest model that fits the memory you have. These picks are re-ranked per tier, so each one uses its budget rather than simply being small.

MemoryTypical hardwareRecommended models
8 GBRTX 4060, RTX 3070, base MacBook AirLlama 3.2 11B Vision Instruct (91.9)
Qwen 2.5 VL 7B Instruct (91.8)
Gemma 4 E4B (89.9)
12 GBRTX 3060 12 GB, RTX 5070Qwen 3.5 14B (93.3)
Gemma 4 12B (92.9)
Llama 3.2 11B Vision Instruct (91.9)
16 GBRTX 5080, RTX 4080, RX 9070 XTMistral Small 3.1 24B (95.5)
Qwen 3.5 14B (93)
Gemma 4 12B (91.5)
24 GBRTX 4090, RTX 3090, RX 7900 XTXGemma 4 27B ⭐ (97.2)
Qwen 3.7 35B-A3B (95.6)
Mistral Small 3.1 24B (95.5)
48 GBRTX 6000 Ada, MacBook Pro M4 Max 48 GBQwen 2.5 VL 72B Instruct (98.5)
Qwen 3.5 72B (97.1)
Gemma 4 27B ⭐ (93.6)
128 GB+Mac Studio, DGX Spark, multi-GPUQwen 2.5 VL 72B Instruct (94.8)
Qwen 3.5 72B (93.3)
Llama 3.2 90B Vision Instruct (91.5)

How this ranking works

Shares the vision hard gate but weights structured-output ability higher (25% coding) and latency higher than general vision, because OCR is typically run in batch over many pages where throughput compounds.

Worked example — Gemma 4 27B ⭐: capability 93.5 × 0.393, quality 92.7 × 0.231, context 100 × 0.15, license 70 × 0.052, accessibility 80 × 0.173 + 3 tag bonus (vision).

Requirements applied: context floor 8,192 tokens (ideal 32,768), quality floor 45, licence weight 0.5, latency weight 0.8. Vision capability is required.

Running OCR and document scanning locally

FAQ

What is the best local LLM for OCR and document scanning?

Gemma 4 27B ⭐, scoring 93.7/100 against this workload's published requirements. 31 models qualified.

What hardware do I need for OCR and document scanning?

A credible answer starts at 8 GB of memory. Larger budgets unlock materially stronger models — the table above lists the best pick at each tier.

How were these models ranked?

Shares the vision hard gate but weights structured-output ability higher (25% coding) and latency higher than general vision, because OCR is typically run in batch over many pages where throughput compounds.

Hardware for This Workload

Related Workloads

Top Pick

Tools

← All workloads | Check your hardware