ProductClank Agents

Zero markup.
You pay providers directly.

OpenClaw is BYOK (Bring Your Own Key) — API costs go straight to the provider. Below is a full breakdown of what each provider charges per 1M tokens. OpenClaw adds no markup.

2
Free options
15+
Under $1/1M tokens
0%
Markup from us
14+
Providers supported
How BYOK pricing works: You add your API key for any provider in your OpenClaw dashboard. When your agent runs, API calls go directly from your instance to that provider. You see the charges on your provider bill — not ours. OpenClaw charges a flat monthly fee ($29–$99/mo) for the platform itself, separate from API usage.

Quickest picks by budget

Free to start
Cloudflare Workers AI, Zhipu GLM-4 Flash

10K neurons/day free (Cloudflare) or fully free (GLM-4 Flash). Great for low-volume agents.

Budget-friendly
Groq, Together AI, Hugging Face

From $0.05/1M input tokens. Groq's LPU hardware gives the fastest inference at the lowest cost.

Best agentic value
MiniMax M2.5, Gemini 3 Flash, GLM-4.7

Strong SWE-bench scores (73–80%) at budget prices. Ideal for complex agentic workflows without premium cost.

Full pricing comparison

Per 1M tokens · Last verified March 2026
Cloudflare Workers AI
Llama 3.3 70BRecommendedFree tier
Free
10K neurons/day free
Free
128K
Fast
Zhipu AI
GLM-4 FlashFastFree
Free
Free
Free
128K
Very Fast
Groq
Llama 3.1 8B InstantFastCheapest paid
$0.05
$0.08
128K
Ultra Fast
Hugging Face
Llama 3.1 8BFast
$0.10
$0.10
128K
Fast
Zhipu AI
GLM-4.7Recommended73.8% SWE
$0.10
$0.30
128K
Fast
Moonshot (Kimi)
Kimi K2Fast
$0.15
$2.50
200K
Fast
Google (Gemini)
Gemini 3 FlashFastBest value flagship78% SWE
$0.15
$0.60
1M
Very Fast
Together AI
Llama 3.1 8B TurboFast
$0.18
$0.18
128K
Very Fast
MiniMax
M2.5 LightningFast
$0.20
$1.10
1M
Very Fast
Mistral AI
Mistral SmallFast
$0.20
$0.60
32K
Fast
MiniMax
MiniMax M2.5RecommendedSOTA agentic80.2% SWE
$0.20
$1.10
1M
Fast
Moonshot (Kimi)
Kimi K2.5Recommended
$0.25
$5.00
200K
Fast
OpenAI
GPT-4.1 MiniFast
$0.40
$1.60
1M
Very Fast
Hugging Face
Llama 3.3 70BRecommendedBest value open-source
$0.40
$0.40
128K
Fast
Groq
Llama 3.3 70BRecommendedBest speed/price
$0.59
$0.79
128K
Ultra Fast
Anthropic
Claude Haiku 4.5Fast
$0.80
$4.00
200K
Very Fast
Together AI
Llama 3.3 70B TurboRecommended
$0.88
$0.88
128K
Fast
Mistral AI
Mistral LargeRecommended
$2.00
$6.00
128K
Fast
xAI (Grok)
Grok 4.1 FastFast
$2.00
$10.00
2M
Fast
OpenAI
GPT-4.1Recommended
$2.00
$8.00
1M
Fast
Google (Gemini)
Gemini 3.1 ProRecommended77% SWE
$2.50
$10.00
2M
Medium
Anthropic
Claude Sonnet 4.6Recommended
$3.00
$15.00
1M
Fast
Together AI
DeepSeek R1Powerful79.8% SWE
$3.00
$7.00
128K
Medium
xAI (Grok)
Grok 4.1Recommended
$5.00
$25.00
2M
Fast
Anthropic
Claude Opus 4.6Powerful80.8% SWE
$15.00
$75.00
1M
Medium
OpenAI
o1Powerful
$15.00
$60.00
200K
Medium
Venice AI
Llama 3.3 70BRecommended
Varies
Subscription-based (~$10/mo unlimited)
Varies
128K
Fast
OpenRouter
Auto (Smart Routing)Recommended
Varies
Varies by model + ~5% markup
Varies
2M
Fast

Prices in USD, approximate. Output tokens are typically 2–5× more expensive than input tokens (Anthropic) or equal (Groq, Together AI). Prices change frequently — always verify at the provider's official pricing page before committing to a provider. Last verified March 2026.

How much will I actually spend?

Light personal use
Daily briefings, reminders, ~50 messages/day
~1M tokens/mo
Budget (Groq)Groq: ~$0.07/mo
Mid (GPT-4.1)GPT-4.1: ~$2.50/mo
PremiumClaude Sonnet: ~$3/mo
Active assistant
Work research, coding help, ~500 messages/day
~10M tokens/mo
Budget (Groq)Groq: ~$0.70/mo
Mid (GPT-4.1)GPT-4.1: ~$25/mo
PremiumClaude Sonnet: ~$30/mo
Heavy automation
Agents, pipelines, 5K+ messages/day
~100M tokens/mo
Budget (Groq)Groq: ~$7/mo
Mid (GPT-4.1)GPT-4.1: ~$250/mo
PremiumClaude Opus: ~$1,500/mo

Estimates assume ~1K input + 500 output tokens per message. Actual costs vary significantly by conversation length and content.

Start with any provider, switch anytime

Not happy with your provider's pricing? Swap to a different one in your dashboard settings — your history, agents, and integrations all stay intact.

Start Free TrialCompare model rankingsAll provider features