Llama 3.2 1B Instruct
edgeAI-generated content marks
The provider reports that this model does not add embedded watermarks or provenance metadata to generated output.
- 1.2B
- 128k
- llama 3
- 🇺🇸 USA
- Sep 2024
Scores
Score per dollar
1296 pts per $/M input
general_score (35) divided by cheapest input price ($0.03/M). Higher is better value. See live pricing.
Related models
Save your hardware and every model page answers the real question: will it run on your machine, and how fast?
Join free - save your rig →Run it locally
Per-quant memory needs and a static "can you run it?" reference - no rig entry required
The reference hardware
22 reference configs, drawn in-house. Scroll for more.
Can you run it? - reference rigs
| Rig | Q4_K_M | Q8_0 |
|---|---|---|
| 4x H100 80GB (320GB) | fast 5161.1t/s | fast 4265.1t/s |
| NVIDIA DGX Station 748GB | fast 3081.2t/s | fast 2546.3t/s |
| 8x RTX 3090 rack (192GB) | fast 2884.7t/s | fast 2383.8t/s |
| 4x RTX 5090 (128GB) | fast 2760.8t/s | fast 2281.5t/s |
| AMD Instinct MI300X (192GB) | fast 2050.9t/s | fast 1694.8t/s |
| 4x RTX 4090 (96GB) | fast 1552.9t/s | fast 1283.3t/s |
| 2x RTX 5090 (64GB) | fast 1380.4t/s | fast 1140.7t/s |
| 2x RTX 3090 (48GB) | fast 721.2t/s | fast 596.0t/s |
| Single RTX 5090 (32GB) | fast 690.2t/s | fast 570.4t/s |
| RTX PRO 6000 Blackwell (96GB) | fast 690.2t/s | fast 570.4t/s |
| Mac Studio M4 Ultra 192GB | fast 458.8t/s | fast 379.2t/s |
| Mac Studio M4 Ultra 512GB | fast 458.8t/s | fast 379.2t/s |
| Single RTX 4090 (24GB) | fast 388.2t/s | fast 320.8t/s |
| MacBook Pro M5 Max 128GB | fast 258.0t/s | fast 213.2t/s |
| Single GTX 1080 Ti (11GB) | fast 186.4t/s | fast 154.1t/s |
| Dual EPYC 9004 + 768GB DDR5-4800 | fast 177.5t/s | fast 146.7t/s |
| DGX Spark 128GB unified | fast 105.2t/s | fast 86.9t/s |
| Ryzen AI Max+ 395 128GB | fast 98.6t/s | fast 81.5t/s |
| Jetson AGX Orin 64GB | fast 78.9t/s | fast 65.2t/s |
| Epyc + 512GB DDR4-3200 + 2x RTX 3090 | fast 78.9t/s | fast 65.2t/s |
| Epyc + 512GB DDR4-2400 + 2x RTX 3090 | fast 59.2t/s | fast 48.9t/s |
| NVIDIA Jetson Orin NX 16GB | fast 39.4t/s | fast 32.6t/s |
Fit tiers use the same will-it-run logic as the rig finder. For comfortable fits, the badge reflects decode speed: fast >=20 t/s, ok 8-20 t/s, slow <8 t/s. t/s is a bandwidth estimate, not a measured benchmark.
Download options
Or run it in the cloud
Live per-provider pricing, throughput and uptime - refreshed about 11 hours ago via OpenRouter. Click a column to sort.
| Provider | Type | Input $/M | Output $/M | Cache $/M | Tok/s | Latency | Uptime | Value |
|---|---|---|---|---|---|---|---|---|
|
Cloudflare
|
API | 0.03 | 0.20 | - | - | - | 100.00% | best uptime |
Default order: throughput among 95%+ uptime providers, then latency; subscriptions last. Sort by any column. Subscription rows show $/mo in the Value column - per-token columns are "-". Affiliate links are marked sponsored / nofollow. Confirm current pricing on the provider's site before committing.
Detailed API pricing page + JSON endpoint →
See who runs Meta in production →
Inference cost over time
Data accumulates from the first daily sync - longer ranges populate over time. Prices come from OpenRouter snapshots, not a historical API.
Tokenstead community
Need a second opinion on this model?
Ask the people running it on real hardware, then post your tok/s and fit notes so the next reader has a data point.
Something wrong or stale? Send a correction.