Cost‑Effective Ollama Cloud Models (2026)
Ollama Cloud provides access to a curated catalog of open‑source LLMs. Below is a quick‑reference table of models that deliver the best performance per dollar.
Cost Ranking (1M input + 1M output tokens)
| Rank | Model | Input ($) | Output ($) | Combined ($) |
|---|---|---|---|---|
| 1 | Nemotron 3 Nano | 0.06 | 0.24 | 0.30 |
| 2 | GPT-OSS 20B | 0.07 | 0.30 | 0.37 |
| 3 | Gemma 4 | 0.14 | 0.40 | 0.54 |
| 4 | Nemotron 3 Super | 0.015 | 0.60 | 0.615 |
| 5 | GLM-5.3 Flash | 0.15 | 0.50 | 0.65 |
| 6 | DeepSeek V4.1 Flash | 0.15 | 0.60 | 0.75 |
| 7 | GPT-OSS 120B | 0.15 | 0.60 | 0.75 |
| 8 | DeepSeek V4 Flash | 0.22 | 0.66 | 0.88 |
| 9 | MiniMax M2.7 | 0.30 | 1.20 | 1.50 |
| 10 | Mistral Large 3 | 0.50 | 1.50 | 2.00 |
| 11 | DeepSeek V4 Pro | 0.66 | 1.98 | 2.64 |
| 12 | MiniMax M3 | 0.60 | 2.40 | 3.00 |
| 13 | Nemotron 3 Ultra | 0.10 | 3.00 | 3.10 |
| 14 | GLM-5.1 | 1.00 | 3.20 | 4.20 |
| 15 | Qwen 3.5 397B | 0.60 | 3.60 | 4.20 |
| 16 | Kimi K2.6 | 0.95 | 4.00 | 4.95 |
| 17 | Kimi K2.7 Code | 0.95 | 4.00 | 4.95 |
| 18 | GLM-5.2 | 1.40 | 4.40 | 5.80 |
| 19 | GLM-5.3 | 1.40 | 4.40 | 5.80 |
| 20 | Kimi K3 | 3.00 | 15.00 | 18.00 |
Parameter‑Size Ranking
| Model | Total Parameters |
|---|---|
| Nemotron 3 Nano | 4B / 30B variants |
| Gemma 4 | 2B–31B variants |
| GPT-OSS 20B | 20B |
| Nemotron 3 Super | 120B |
| GPT-OSS 120B | 120B |
| MiniMax M2.7 | ~229B |
| DeepSeek V4 Flash | 304B |
| GLM-5.3 Flash | 321B |
| Qwen 3.5 | 397B |
| MiniMax M3 | ~428B |
| Nemotron 3 Ultra | 550B |
| Mistral Large 3 | 675B |
| GLM-5.3 | 753B |
| GLM-5.2 | 756B |
| DeepSeek V4.1 Flash | 763B |
| Kimi K2.6 | 1.04T |
| Kimi K2.7 Code | 1.04T |
| DeepSeek V4 Pro | 1.65T |
| Kimi K3 | 2.81T |
These rankings help you pick a model that balances cost, parameter size, and performance for your workload.