Models · list price

What your credits buy.

Every model you can run on Daslab, and what a million tokens of it costs. These are the provider’s own list prices. We pass them through and take nothing on top, so the number below is the number that leaves your balance.

80 models · 20 vendors · checked 2026-08-20
How well do they work? Reliability and cost per task →
Anthropic
Model Context First token Tokens / s Input / 1M Output / 1M
Claude 3 Haiku vision # 200K — — $0.25 $1.25
Claude 3.5 Haiku vision # 200K — — $0.25 $1.25
Claude Haiku 4.5 vision # 200K 1.1s 122 $1.00 $5.00
Claude 3.5 Sonnet v2 vision # 200K — — $3.00 $15.00
Claude 3.7 Sonnet vision # 200K — — $3.00 $15.00
Claude Sonnet 4 vision # 200K — — $3.00 $15.00
Claude Opus 4.8 vision # 1M 2.1s 74 $5.00 $25.00
Claude Opus 5 vision reasoning # 1M — — $5.00 $25.00
OpenAI
Model Context First token Tokens / s Input / 1M Output / 1M
gpt-oss-20b reasoning # 131K — — $0.03 $0.14
gpt-oss-120b reasoning # 131K — — $0.037 $0.17
GPT-5.2-Codex vision reasoning # 400K — — $1.75 $14.00
GPT-5.6 (Sol alias) vision # 1.1M — — $5.00 $30.00
GPT-5.4 Image 2 vision reasoning no tools # 272K — — $8.00 $15.00
GPT-5.5 Pro vision reasoning # 1.1M — — $30.00 $180.00
Google
Model Context First token Tokens / s Input / 1M Output / 1M
Gemma 4 26B A4B vision reasoning # 262K — — $0.07 $0.34
Gemma 4 31B vision reasoning # 262K — — $0.10 $0.34
Gemini 3.1 Flash Lite vision reasoning # 1M — — $0.25 $1.50
Gemini 3.5 Flash Lite vision reasoning # 1M — — $0.30 $2.50
Gemini 3.7 Flash vision reasoning # 1M 3.1s 316 $0.375 $1.88
Gemini 3.6 Flash vision reasoning # 1M — — $0.75 $3.75
Gemini 3.5 Flash vision reasoning # 1M 5.0s 180 $1.50 $9.00
Gemini 3.1 Pro Preview vision reasoning # 1M — — $2.00 $12.00
Meta
Model Context First token Tokens / s Input / 1M Output / 1M
Llama 3.1 8B Turbo # 131K — — $0.18 $0.18
Llama 3 8B Instruct # 8K — — $0.20 $0.20
Llama 4 Sonnet 17B vision # 41K — — $0.60 $2.40
Llama 3.1 70B Turbo # 131K — — $0.88 $0.88
Llama 3.3 70B Turbo # 131K — — $0.88 $0.88
Muse Spark 1.1 vision reasoning # 1M — — $1.25 $4.25
Muse Spark 1.2 vision reasoning # 1M — — $1.25 $4.25
Llama 3 70B Instruct # 8K — — $2.65 $3.50
Mistral
Model Context First token Tokens / s Input / 1M Output / 1M
Mistral Nemo # 131K — — $0.019 $0.03
Mixtral 8x7B # 32K — — $0.24 $0.24
Mistral Large # 128K — — $2.00 $6.00
DeepSeek
Model Context First token Tokens / s Input / 1M Output / 1M
DeepSeek V4 Flash 0731 reasoning # 1.3M 1.7s 68 $0.09 $0.18
DeepSeek V4 Flash reasoning # 1M — — $0.14 $0.28
DeepSeek V4 Pro 0813 reasoning # 1M — — $1.22 $2.44
DeepSeek V3 # 131K — — $1.25 $1.25
DeepSeek V4 Pro 0423 reasoning # 1M — — $1.30 $2.60
DeepSeek R1 # 164K — — $3.00 $7.00
xAI
Model Context First token Tokens / s Input / 1M Output / 1M
Grok Build 0.1 vision reasoning # 256K — — $1.00 $2.00
Grok 4.20 vision reasoning # 2M — — $1.25 $2.50
Grok 4.3 vision reasoning # 1M — — $1.25 $2.50
Grok 4.5 vision reasoning # 500K — — $2.00 $6.00
Grok 4.6 vision reasoning # 500K — — $2.00 $6.00
Alibaba (Qwen)
Model Context First token Tokens / s Input / 1M Output / 1M
Qwen3.7 Flash vision reasoning # 1M — — $0.03 $0.13
Qwen3.5-35B-A3B vision reasoning # 262K — — $0.14 $1.00
Qwen3.6 35B A3B vision reasoning # 262K — — $0.15 $1.00
Qwen3 Coder 480B A35B # 262K — — $0.30 $1.00
Qwen3 Max # 262K — — $0.78 $3.90
Qwen 2.5 Coder 32B # 33K — — $0.80 $0.80
Qwen 2.5 72B Turbo # 131K — — $1.20 $1.20
QwQ 32B Preview # 33K — — $1.20 $1.20
Qwen3.7 Max reasoning # 1M — — $1.48 $4.42
Qwen3.8 Max vision reasoning # 1M — — $2.00 $6.00
Azure
Model Context First token Tokens / s Input / 1M Output / 1M
GPT-5 Nano vision reasoning # 400K — — $0.05 $0.40
GPT-5.6 Luna vision reasoning # 1.1M — — $1.00 $6.00
Claude Sonnet 5 vision reasoning # 1M 3.5s 128 $2.00 $10.00
GPT-5.6 Terra vision reasoning # 1.1M — — $2.50 $15.00
GPT-5.6 Sol vision reasoning # 1.1M — — $5.00 $30.00
Moonshot AI
Model Context First token Tokens / s Input / 1M Output / 1M
Kimi K2 Thinking reasoning # 262K — — $0.60 $2.50
Kimi K2.6 vision reasoning # 262K — — $0.60 $3.41
Kimi K2.7 Code vision reasoning # 262K 774ms 110 $0.73 $3.50
Kimi K3 vision reasoning # 1M 4.9s 44 $3.00 $15.00
Z.ai
Model Context First token Tokens / s Input / 1M Output / 1M
GLM 4.5 Air # 131K — — $0.20 $0.20
GLM 5.2 reasoning # 1M — — $0.284 $0.893
GLM 4.6V vision reasoning # 131K — — $0.30 $0.90
AI21
Model Context First token Tokens / s Input / 1M Output / 1M
Jurassic-2 Mid # 8K — — $0.35 $1.15
Jurassic-2 Ultra # 8K — — $0.90 $4.50
Amazon
Model Context First token Tokens / s Input / 1M Output / 1M
Nova Lite vision # 300K — — $0.06 $0.24
Nova Pro vision # 300K — — $0.80 $3.20
Cohere
Model Context First token Tokens / s Input / 1M Output / 1M
Command R # 128K — — $0.50 $1.50
Command R+ # 128K — — $0.50 $2.50
Thinking Machines
Model Context First token Tokens / s Input / 1M Output / 1M
Inkling Small vision # 256K — — $0.50 $1.50
Inkling vision # 256K — — $3.00 $7.00
Xiaomi
Model Context First token Tokens / s Input / 1M Output / 1M
MiMo-V2.5-Pro reasoning # 1.1M — — $0.435 $0.87
MiMo-V2-Pro no tools # — — — $2.00 $6.00
Claude-fable-5
Model Context First token Tokens / s Input / 1M Output / 1M
Claude Fable 5 vision reasoning # 1M — — $10.00 $50.00
MiniMax
Model Context First token Tokens / s Input / 1M Output / 1M
MiniMax M3 vision reasoning # 1M — — $0.30 $1.20
StepFun
Model Context First token Tokens / s Input / 1M Output / 1M
Step 3.5 Flash reasoning # 262K — — $0.10 $0.30
Tencent
Model Context First token Tokens / s Input / 1M Output / 1M
Hy3 preview reasoning # 262K — — $0.063 $0.21

How to read this. Prices are per million tokens, in US dollars, at the provider’s list price. Running an agent spends your balance at exactly these rates: first the usage allowance included in your plan, then any top-up credit. A typical run costs $0.25 to $0.75 depending on the model and the size of the job.

We take no markup. Model usage is passed through at cost. Every price here is also published by OpenRouter, models.dev and LiteLLM’s price file, and we reconcile against all three twice a day.

Bring your own key instead. Any model you already pay for elsewhere can be connected directly, in which case it never touches your Daslab balance at all. Prices above apply only to usage you buy through us.