OpenAI GPT Latest
agent
~openai/gpt-latest
1,050,000
Per 1K tokens: input 7.75 points, output 46.51 points
Usage-based, prepaid points
Top up points and pay by model price.
PacketLab uses one points wallet. Top up first, then pay by each model's input and output rate; failed, limited, and service-error calls are not charged.
Model pricing
Current public model pricing. Points are platform billing units, not currency; input and output are priced separately per 1K tokens.
ModelModel IDContextPricing
AI21: Jamba Large 1.7
agent
ai21/jamba-large-1.7
256,000
Per 1K tokens: input 3.3 points, output 13.2 points
AionLabs: Aion-1.0
Reasoning
aion-labs/aion-1.0
131,072
Per 1K tokens: input 6.2 points, output 12.4 points
AionLabs: Aion-1.0-Mini
Reasoning
aion-labs/aion-1.0-mini
131,072
Per 1K tokens: input 1.16 points, output 2.31 points
AionLabs: Aion-2.0
chinese
aion-labs/aion-2.0
131,072
Per 1K tokens: input 1.32 points, output 2.64 points
AionLabs: Aion-RP 1.0 (8B)
General
aion-labs/aion-rp-llama-3.1-8b
32,768
Per 1K tokens: input 1.32 points, output 2.64 points
AlfredPros: CodeLLaMa 7B Instruct Solidity
Coding
alfredpros/codellama-7b-instruct-solidity
4,096
Per 1K tokens: input 1.44 points, output 2.16 points
AllenAI: Olmo 3 32B Think
Reasoning
allenai/olmo-3-32b-think
65,536
Per 1K tokens: input 0.27 points, output 0.9 points
AllenAI: Olmo 3.1 32B Instruct
agent
allenai/olmo-3.1-32b-instruct
65,536
Per 1K tokens: input 0.36 points, output 1.08 points
Amazon: Nova 2 Lite
Reasoning
amazon/nova-2-lite-v1
1,000,000
Per 1K tokens: input 0.5 points, output 4.13 points
Amazon: Nova Lite 1.0
agent
amazon/nova-lite-v1
300,000
Per 1K tokens: input 0.11 points, output 0.44 points
Amazon: Nova Micro 1.0
agent
amazon/nova-micro-v1
128,000
Per 1K tokens: input 0.07 points, output 0.27 points
Amazon: Nova Premier 1.0
Reasoning
amazon/nova-premier-v1
1,000,000
Per 1K tokens: input 3.88 points, output 19.38 points
Amazon: Nova Pro 1.0
agent
amazon/nova-pro-v1
300,000
Per 1K tokens: input 1.32 points, output 5.28 points
Anthropic Claude Haiku Latest
agent
~anthropic/claude-haiku-latest
200,000
Per 1K tokens: input 1.65 points, output 8.25 points
Anthropic Claude Sonnet Latest
agent
~anthropic/claude-sonnet-latest
1,000,000
Per 1K tokens: input 4.66 points, output 23.26 points
Anthropic: Claude 3 Haiku
agent
anthropic/claude-3-haiku
200,000
Per 1K tokens: input 0.45 points, output 2.25 points
Anthropic: Claude 3.5 Haiku
agent
anthropic/claude-3.5-haiku
200,000
Per 1K tokens: input 1.32 points, output 6.6 points
Anthropic: Claude 3.7 Sonnet
Reasoning
anthropic/claude-3.7-sonnet
200,000
Per 1K tokens: input 4.66 points, output 23.26 points
Anthropic: Claude 3.7 Sonnet (thinking)
Reasoning
anthropic/claude-3.7-sonnet:thinking
200,000
Per 1K tokens: input 4.66 points, output 23.26 points
Anthropic: Claude Haiku 4.5
agent
anthropic/claude-haiku-4.5
200,000
Per 1K tokens: input 1.65 points, output 8.25 points
Anthropic: Claude Opus 4
agent
anthropic/claude-opus-4
200,000
Per 1K tokens: input 21.76 points, output 108.75 points
Anthropic: Claude Opus 4.1
Reasoning
anthropic/claude-opus-4.1
200,000
Per 1K tokens: input 21.76 points, output 108.75 points
Anthropic: Claude Opus 4.5
Reasoning
anthropic/claude-opus-4.5
200,000
Per 1K tokens: input 7.75 points, output 38.75 points
No pre-charge
Adding models grants access; completed responses consume points by usage.
Token budgets
Each API token can have budgets, model scopes, and IP allowlists.
Clear ledger
Usage logs, top-ups, and wallet ledger stay traceable.