tokenstat
tokenstat

Model catalog

Models and list rates

Published API list rates (USD per million tokens), who lists each model, and whether it can do tools, vision, or reasoning. Recommended order is newer first; among similarly new models, larger context wins. Free/$0 rows stay hidden unless you ask. Not a bill: measure real sessions with the CLI.

3,231 models in the merged catalog · as of . Sources: LiteLLM, OpenRouter, models.dev. · Providers · Plans · Benchmarks

Showing 48 of 2,670 matches · page 53 of 56 · 561 free/$0 rows hidden (show them)

ModelInputOutputContextOffers
Llama 3.2 90B Vision Instructwatsonx/meta-llama/llama-3-2-90b-vision-instruct
$2LiteLLM$2N/A1
Llama3 70B Instructnovita/meta-llama/llama-3-70b-instruct
$0.51LiteLLM$0.74N/A1
Llama3 70B Instructopenrouter/meta-llama/llama-3-70b-instruct
$0.59LiteLLM$0.79N/A1
Llama3 70B Instructreplicate/meta/llama-3-70b-instruct
$0.65LiteLLM$2.75N/A1
Meta-Llama-3-70B-Instructanyscale/meta-llama/Meta-Llama-3-70B-Instruct
$1LiteLLM$1N/A1
Meta-Llama-3-70B-Instructazure_ai/Meta-Llama-3-70B-Instruct
$1.10LiteLLM$0.37N/A1
Meta-Llama-3-70B-Instructhyperbolic/meta-llama/Meta-Llama-3-70B-Instruct
$0.12LiteLLM$0.3N/A1
Meta-Llama-3-70B-InstructMeta-Llama-3-70B-Instruct
$1LiteLLM$1N/A1
CSM 1Bsesame/csm-1b
$7OpenRouter$04K1
Orpheus 3Bcanopylabs/orpheus-3b-0.1-ft
$7OpenRouter$04K1
Deepseek R1 0528 Distill Qwen3 8baccounts/fireworks/models/deepseek-r1-0528-distill-qwen3-8b
$0.2LiteLLM$0.2N/A3
Ft:Gpt 3.5 Turboft:gpt-3.5-turbo
$3LiteLLM$6N/A1
Ft:Gpt 3.5 Turbo 0125ft:gpt-3.5-turbo-0125
$3LiteLLM$6N/A1
Ft:Gpt 3.5 Turbo 0613ft:gpt-3.5-turbo-0613
$3LiteLLM$6N/A1
Ft:Gpt 3.5 Turbo 1106ft:gpt-3.5-turbo-1106
$3LiteLLM$6N/A1
Ft:Gpt 4 0613ft:gpt-4-0613
$30LiteLLM$60N/A1
Gpt 3.5 Turbo Instruct 0914azure/gpt-3.5-turbo-instruct-0914
$1.50LiteLLM$2N/A2
Gpt 4 0125 Previewazure/gpt-4-0125-preview
$10LiteLLM$30N/A2
Gpt 4 0314gpt-4-0314
$30LiteLLM$60N/A1
Gpt 4 0613azure/gpt-4-0613
$30LiteLLM$60N/A2
Gpt 4 1106 Previewazure/gpt-4-1106-preview
$10LiteLLM$30N/A2
Gpt 4 32kazure/gpt-4-32k
$60LiteLLM$120N/A2
Gpt 4 32k 0613azure/gpt-4-32k-0613
$60LiteLLM$120N/A2
Gpt 4 Turbo 2024 04 09azure/gpt-4-turbo-2024-04-09
$10LiteLLM$30N/A2
Gpt 4 Turbo Vision Previewazure/gpt-4-turbo-vision-preview
$10LiteLLM$30N/A2
Gpt 4.5 Previewazure/gpt-4.5-preview
$75LiteLLM$150N/A2
Anthropic: Claude 3 Haikuvertex_ai/claude-3-haiku@20240307
$0.25LiteLLM$1.25N/A1
Anthropic: Claude Opus 3vertex_ai/claude-3-opus@20240229
$15LiteLLM$75N/A1
Anthropic.Claude 3 Haiku 20240307 V1:0bedrock/us-gov-east-1/anthropic.claude-3-haiku-20240307-v1:0
$0.25LiteLLM$1.25N/A8
Anthropic.Claude 3 Sonnet 20240229 V1:0anthropic.claude-3-sonnet-20240229-v1:0
$3LiteLLM$15N/A5
Claude 3 Opus 20240229 V1:0claude-3-opus-20240229-v1:0
$15LiteLLM$75N/A4
Claude 3 Sonnet@20240229vertex_ai/claude-3-sonnet@20240229
$3LiteLLM$15N/A2
Google: Flan T5 Xl 3bgoogle/flan-t5-xl-3b
$0.6LiteLLM$0.6N/A3
Google: Gemma 7b Itgoogle/gemma-7b-it
$0.05LiteLLM$0.08N/A6
Ministral 3 8B 2512mistral/ministral-8b-2512
scores
$0.15LiteLLM$0.15N/A1
Mistral 7b Instructmistralai/mistral-7b-instruct
$0.13LiteLLM$0.13N/A1
Mistral 7b V0.1mistralai/mistral-7b-v0.1
$0.05LiteLLM$0.25N/A3
Mistral: Ministral 3 14b 2512mistral/ministral-3-14b-2512
$0.2LiteLLM$0.2N/A2
Mistral: Ministral 3 3b 2512mistral/ministral-3-3b-2512
$0.1LiteLLM$0.1N/A2
Mistral: Ministral 3 8b 2512mistral/ministral-3-8b-2512
$0.15LiteLLM$0.15N/A2
Mistral 7b Instructperplexity/mistral-7b-instruct
$0.07LiteLLM$0.28N/A1
Pplx 7b Chatperplexity/pplx-7b-chat
$0.07LiteLLM$0.28N/A2
Pplx 7b Onlineperplexity/pplx-7b-online
$0LiteLLM$0.28N/A2
Code Llama 7baccounts/fireworks/models/code-llama-7b
$0.2LiteLLM$0.2N/A3
Code Llama 7b Instructaccounts/fireworks/models/code-llama-7b-instruct
$0.2LiteLLM$0.2N/A3
Code Llama 7b Pythonaccounts/fireworks/models/code-llama-7b-python
$0.2LiteLLM$0.2N/A3
Code Qwen 1p5 7baccounts/fireworks/models/code-qwen-1p5-7b
$0.2LiteLLM$0.2N/A3
Codegemma 7baccounts/fireworks/models/codegemma-7b
$0.2LiteLLM$0.2N/A3

Price this against your own usage

List rates are only half the story. The CLI reads the session logs your coding harnesses already write and shows tokens, cache share, and usage value on your machine.

$ curl -fsSL https://tokenstat.ai/install.sh | bash