modelgrep

Small & Fast Meta Models

Quick answer · Updated June 2026

The small, fast Meta model is Llama 3.1 8B Instruct — the efficient tier at 240 tokens/sec and $0.020 per million input tokens. It trades a few points of raw intelligence for speed and cost, the right call for high-volume, latency-sensitive work. Llama 3.2 3B Instruct (free) (149 t/s) and Llama 3.2 3B Instruct (149 t/s) round out the top three.

240 t/sSpeed
11.8Intelligence
$0.020Input /M
131KContext

Compact, efficient models — the small/mini/flash/haiku tier — ranked by output speed. These trade a little raw intelligence for low cost and high throughput, which is the right tradeoff for chat, classification, extraction and other high-volume work.

  1. 1M
    llama-3.1-8b-instruct
    ToolsJSON11.8 intel · $0.020/M · 136ms ttft
    240 t/s
    Speed
  2. 2M
    llama-3.2-3b-instruct:free
    Free/M · 228ms ttft · 131K ctx
    149 t/s
    Speed
  3. 3M
    llama-3.2-3b-instruct
    $0.051/M · 228ms ttft · 131K ctx
    149 t/s
    Speed
  4. 4M
    llama-4-scout
    ToolsJSONVision13.5 intel · $0.100/M · 355ms ttft
    104 t/s
    Speed
  5. 5M
    llama-3-8b-instruct
    6.4 intel · $0.140/M · 798ms ttft
    64 t/s
    Speed
  6. 6M
    llama-4-maverick
    ToolsJSONVision18.4 intel · $0.150/M · 294ms ttft
    61 t/s
    Speed
  7. 7M
    llama-3.2-1b-instruct
    6.3 intel · $0.027/M · 316ms ttft
    60 t/s
    Speed
  8. 8M
    llama-3.2-11b-vision-instruct
    JSONVision8.7 intel · $0.345/M · 195ms ttft
    33 t/s
    Speed
  9. 9M
    llama-guard-4-12b
    JSONVision$0.180/M · 118ms ttft · 164K ctx
    17 t/s
    Speed

Frequently asked

What is the smallest, fastest Meta model?

The small, fast Meta model is Llama 3.1 8B Instruct — the efficient tier at 240 tokens/sec and $0.020 per million input tokens. It trades a few points of raw intelligence for speed and cost, the right call for high-volume, latency-sensitive work. Llama 3.2 3B Instruct (free) (149 t/s) and Llama 3.2 3B Instruct (149 t/s) round out the top three.

What's a good alternative to Llama 3.1 8B Instruct?

Llama 3.2 3B Instruct (free) (149 t/s) is the closest alternative on this metric, followed by Llama 3.2 3B Instruct (149 t/s). See the full ranking above for the tradeoffs.

How many Meta models are there?

modelgrep tracks 13 Meta models with live benchmarks, speed, latency and per-provider pricing, led on intelligence by Llama 4 Maverick. 9 of them qualify for this ranking.

More Meta rankings

All rankings