Calculate your Ollama Cloud bill from published per-token rates, see which Ollama credit plan fits, and compare it with a flat Plugsky plan at the same volume.
Prices checked 19 Nov 2026
Off-peak applies to DeepSeek models only.
Token volume
—
Ollama Cloud monthly cost
—
—
Ollama Cloud plan fit
—
Plugsky flat plan at this volume
—
—
Estimates use list prices and your averages. Cached input is billed at the model’s cached-input rate. The comparison assumes the same token volume on Plugsky; model quality and capabilities differ. Plan concurrency also differs: 3 concurrent requests on Pay as you go and Pro, 10 on Max and Team.
Per million tokens, from Ollama’s public pricing table on 19 November 2026. Off-peak rates apply outside the 12:00–18:00 UTC weekday window and all day at weekends.
| Model | Input | Cached input | Output | Off-peak input / output |
|---|---|---|---|---|
| gpt-oss:120b | $0.15 | $0.014 | $0.60 | — |
| gpt-oss:20b | $0.07 | $0.035 | $0.30 | — |
| deepseek-v4.1-flash | $0.30 | $0.006 | $1.20 | $0.15 / $0.60 |
| deepseek-v4-pro | $1.32 | $0.044 | $3.96 | $0.66 / $1.98 |
| gemma4 | $0.14 | $0.05 | $0.40 | — |
| glm-5.3-flash | $0.15 | $0.03 | $0.50 | — |
| minimax-m3.1-flash | $0.15 | $0.015 | $0.60 | — |
| mimo-v2.6-flash | $0.14 | $0.0028 | $0.28 | — |
| mistral-large-4 | $0.70 | $0.07 | $2.10 | — |
| nemotron-3-super | $0.015 | $0.015 | $0.60 | — |
| nemotron-3-ultra | $0.10 | $0.10 | $3.00 | — |
| kimi-k3.1-pro | $2.00 | $0.20 | $10.00 | — |
| kimi-k3 | $3.00 | $0.30 | $15.00 | — |
Ollama Cloud has a $0 pay-as-you-go base and bills per token. Paid plans pre-buy usage credits at a discount:
| Plan | Price | Included usage | Concurrency | Best for |
|---|---|---|---|---|
| Pay as you go | $0 base | None; pay per token | 3 | Trying cloud models or low volume |
| Pro | $20/mo or $200/yr | $60 credits/month | 3 | Regular use up to about $60/month |
| Max | $100/mo | $300 credits/month | 10 | Parallel agents and heavier use |
| Team | $500/mo | $1,000 shared credits/month | 10 | Teams with shared billing |
| Enterprise | Custom | Volume pricing | Custom | Budgets, access controls, support |
Included credits do not roll over. Ollama sends an email reminder at 90% of included monthly usage. On annual plans, the included amount still refreshes monthly.
If the calculator does not run, use this table. It assumes an 80/20 input-output split and no cached input. Ollama figures are raw pay-as-you-go usage; at 120M, Pro’s $60 credit reduces the effective bill to $20.
| Model | 12M tokens/mo (10M in + 2M out) | 60M tokens/mo (50M in + 10M out) | 120M tokens/mo (100M in + 20M out) |
|---|---|---|---|
| gpt-oss:120b | $2.70 | $13.50 | $27.00 |
| gpt-oss:20b | $1.30 | $6.50 | $13.00 |
| deepseek-v4.1-flash (off-peak) | $2.70 | $13.50 | $27.00 |
| gemma4 | $2.20 | $11.00 | $22.00 |
| kimi-k3 | $60.00 | $300.00 | $600.00 |
| Plugsky flat plan | $8 Hobby | $8 Hobby | $20 Starter |
Ollama Cloud is cheaper at low volume because you pay only for what you use. A flat plan wins when its monthly fee is below the effective Ollama bill after plan credits:
These are arithmetic comparisons at GPT-OSS 120B rates, not quality comparisons. Plan credits change the answer: Pro’s $60 credit makes Ollama cheaper than raw token math suggests between roughly $20 and $60 of monthly usage. If your workload needs a model that only one provider offers, compare the price of that model.
Ollama Cloud is pay-as-you-go per token. GPT-OSS 120B is $0.15 input and $0.60 output per million tokens; GPT-OSS 20B is $0.07 and $0.30. Pro costs $20 per month and includes $60 of usage credits; Max costs $100 and includes $300.
There is no permanent free cloud tier. The pay-as-you-go plan has a $0 base fee but bills per token. Running models on your own hardware with Ollama is free and unlimited, subject to your GPU capacity.
It depends on volume and model. At GPT-OSS 120B rates and an 80/20 input-output mix, 12 million tokens per month costs about $2.70 on Ollama Cloud versus $8 on Plugsky Hobby. At 60 million tokens, Ollama is about $13.50 versus $8 on Hobby. At 120 million, Ollama is about $27 versus $20 on Starter.
Ollama states that prompt and response data is never logged or trained on. Models are hosted primarily in the United States, with Europe used for additional capacity. Read the current privacy policy and cloud FAQ before sending regulated data.
Pay-as-you-go and Pro include 3 concurrent requests; Max and Team include 10. Requests beyond the limit are queued and processed when a slot opens, and can be rejected if the queue is full.
No. Ollama’s cloud documentation states that cloud models currently do not support structured outputs. Test JSON-mode and schema-constrained workflows locally or on another provider before relying on them.
Yes. Point the OpenAI SDK at https://ollama.com/v1 with an Ollama API key. The local server uses http://localhost:11434/v1. Not every OpenAI parameter is supported, so test tool calling and response formats.
Plugsky runs 39 models behind one OpenAI-compatible endpoint. Hobby is $8/month with 2M tokens/day; Starter is $20 with 5M tokens/day; Builder and Scale are fair-use plans with higher concurrency and context.
See Plugsky plans → · General LLM cost calculator → · Cheapest LLM APIs →
Reviewed 19 November 2026. Ollama Cloud prices, plan credits and limits were checked against ollama.com/pricing and docs.ollama.com/cloud on that date. Prices change; confirm before you commit.