OLLAMA CLOUD, UNLEASHED

Model atlas

One card per cloud model on Ollama’s live list. Prices come from the pricing page, with off-peak rows where a model has them.

Verified 12 Oct 2026

Sources: ollama.com/pricing, ollama.com/api/tags (live probe), ollama.com/api/show (live probe)

How to read the cards

  • Prices are dollars per million tokens: fresh input, cached input and output.
  • Off-peak prices apply outside 12:00 to 18:00 UTC on weekdays, and all day at weekends. Only models with an off-peak row get them.
  • “One typical call” is illustrative, not a measurement. It is 10,000 input tokens, 80% of them cached, and 1,000 output tokens. Use it to compare prices, not to predict your bill.
  • Context and capabilities come from our probes of the live API. A model we have not probed says so.

The live list

The live list held 18 cloud models on 12 Oct 2026. Seventeen have a price row. mistral-large-4 is in the list with no price yet, so its card says so.

Sort

Showing every model in alphabetical order.

  1. deepseek-v4-pro

    Live list name: deepseek-v4-pro:0813

    Off-peak row
    Fresh input
    $1.32
    Cached input
    $0.044
    Output
    $3.96
    One typical call
    $0.006952

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Off-peak

    $0.66 fresh input, $0.022 cached, $1.98 output per million tokens. The same illustrative call costs $0.003476 off-peak.

    Off-peak applies outside 12:00 to 18:00 UTC on weekdays, and all day at weekends.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  2. deepseek-v4.1-flash

    Off-peak row
    Fresh input
    $0.30
    Cached input
    $0.006
    Output
    $1.20
    One typical call
    $0.001848

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Off-peak

    $0.15 fresh input, $0.003 cached, $0.60 output per million tokens. The same illustrative call costs $0.000924 off-peak.

    Off-peak applies outside 12:00 to 18:00 UTC on weekdays, and all day at weekends.

    Context: 1,048,576 tokens (probed)

    • completion
    • thinking
    • tools
    • vision

    Our test

    • Per-call test, round 2 (30 Sep 2026, 15 calls per model): 0.140 session points per call, against 0.093 for glm-5.3-flash.
    • Round 1 (27 Aug 2026) tested a different model, deepseek-v4-flash:0731: 0.015 against 0.030 for glm-5.3-flash.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  3. gemma4

    Live list name: gemma4:31b

    Fresh input
    $0.14
    Cached input
    $0.05
    Output
    $0.40
    One typical call
    $0.001080

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context: 262,144 tokens (probed)

    • completion
    • thinking
    • tools
    • vision

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  4. glm-5.2

    Fresh input
    $1.40
    Cached input
    $0.26
    Output
    $4.40
    One typical call
    $0.009280

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  5. glm-5.3

    Fresh input
    $1.40
    Cached input
    $0.26
    Output
    $4.40
    One typical call
    $0.009280

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  6. glm-5.3-flash

    Fresh input
    $0.15
    Cached input
    $0.03
    Output
    $0.50
    One typical call
    $0.001040

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context: 1,048,576 tokens (probed)

    • completion
    • thinking
    • tools
    • vision

    Our test

    • Per-call test, round 1 (27 Aug 2026, 20 calls per model): 0.030 session points per call, against 0.015 for deepseek-v4-flash:0731.
    • Per-call test, round 2 (30 Sep 2026, 15 calls per model): 0.093 session points per call, against 0.140 for deepseek-v4.1-flash.
    • Live probe: with max_tokens set to 5 the reply had empty content and finish_reason length. Reasoning used the whole budget.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  7. gpt-oss:120b

    Fresh input
    $0.15
    Cached input
    $0.014
    Output
    $0.60
    One typical call
    $0.001012

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context: 131,072 tokens (probed)

    • completion
    • thinking
    • tools

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  8. gpt-oss:20b

    Fresh input
    $0.07
    Cached input
    $0.035
    Output
    $0.30
    One typical call
    $0.000720

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  9. kimi-k2.6

    Fresh input
    $0.95
    Cached input
    $0.16
    Output
    $4.00
    One typical call
    $0.007180

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  10. kimi-k2.7-code

    Fresh input
    $0.95
    Cached input
    $0.19
    Output
    $4.00
    One typical call
    $0.007420

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  11. kimi-k3

    Fresh input
    $3.00
    Cached input
    $0.30
    Output
    $15.00
    One typical call
    $0.023400

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context: 1,048,576 tokens (probed)

    • completion
    • thinking
    • tools
    • vision

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  12. minimax-m2.7

    Fresh input
    $0.30
    Cached input
    $0.06
    Output
    $1.20
    One typical call
    $0.002280

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  13. minimax-m3

    Fresh input
    $0.60
    Cached input
    $0.12
    Output
    $2.40
    One typical call
    $0.004560

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context: 512,000 tokens (probed)

    • completion
    • thinking
    • tools
    • vision

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  14. mistral-large-3

    Live list name: mistral-large-3:675b

    Fresh input
    $0.50
    Cached input
    None listed
    Output
    $1.50
    One typical call
    $0.006500

    Per million tokens, standard rates. No cached price is listed, so the illustrative call prices all 10,000 input tokens at the fresh rate (an upper bound).

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  15. mistral-large-4

    No price listed yet.

    Context: 1,048,576 tokens (probed)

    • completion
    • thinking
    • tools
    • vision

    No price row on 12 Oct 2026. Source: ollama.com/pricing

  16. nemotron-3-nano

    Live list name: nemotron-3-nano:30b

    Fresh input
    $0.06
    Cached input
    None listed
    Output
    $0.24
    One typical call
    $0.000840

    Per million tokens, standard rates. No cached price is listed, so the illustrative call prices all 10,000 input tokens at the fresh rate (an upper bound).

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  17. nemotron-3-super

    Fresh input
    $0.015
    Cached input
    $0.015
    Output
    $0.60
    One typical call
    $0.000750

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context: 262,144 tokens (probed)

    • completion
    • thinking
    • tools

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

  18. nemotron-3-ultra

    Fresh input
    $0.10
    Cached input
    $0.10
    Output
    $3.00
    One typical call
    $0.004000

    Per million tokens, standard rates. Illustrative call: 10,000 input tokens, 8,000 of them cached, 1,000 output.

    Context and capabilities: not probed yet.

    Prices verified 12 Oct 2026. Source: ollama.com/pricing

Before you choose

Prices and per-call costs are only half the story. Re-test a model before you commit to it. In our tests the cheapest model changed within one model generation.

Read Cheaper models for how the prices work, and Our tests for the raw numbers behind the cards.

Check the real numbers

Paste an Ollama Cloud key on the home page. Exact monthly dollars on new plans, estimates on old ones, and nothing is stored.

Check your own usage