O OpenAI API

OpenAI API Pricing 2026

GPT-4o, GPT-4o mini, o1, and o1-mini — complete token pricing, batch discounts, and fine-tuning costs.

Updated May 2026
OpenAI Anthropic Google AWS Bedrock Azure OpenAI

Model Pricing

All prices are per 1 million tokens. Output tokens are billed separately and cost more than input tokens — typically 4–5× more.

Model Input / 1M tokens Output / 1M tokens Context window Tier
GPT-4o $2.50 $10.00 128K Production
GPT-4o mini $0.15 $0.60 128K Budget
o1 $15.00 $60.00 200K Reasoning
o1-mini $3.00 $12.00 128K Reasoning
o3-mini $1.10 $4.40 200K Reasoning
text-embedding-3-large $0.13 8K Embeddings

Batch API Discounts

OpenAI's Batch API delivers 50% cost reduction for non-real-time workloads. Requests are processed within 24 hours. Ideal for document classification, evaluation pipelines, and bulk analysis.

Model Batch input / 1M Batch output / 1M Savings vs. standard
GPT-4o $1.25 $5.00 50% off
GPT-4o mini $0.075 $0.30 50% off

Fine-Tuning Costs

Model Training / 1M tokens Inference input / 1M Inference output / 1M
GPT-4o fine-tune $25.00 $3.75 $15.00
GPT-4o mini fine-tune $3.00 $0.30 $1.20
Watch out: Fine-tuned model storage costs $0.25/M tokens/month. A low-utilization fine-tune can cost more to host than it saves in inference quality.

Best For

Production chat & agents

GPT-4o's combination of capability, speed, and wide integration support makes it the default for customer-facing AI features.

High-volume classification

GPT-4o mini at $0.15/M input is the cheapest capable model for bulk classification, tagging, and routing tasks.

Complex multi-step reasoning

o1 and o3-mini excel at code generation, math, and tasks that benefit from extended thinking time over raw throughput.

Rate Limits & Tiers

OpenAI uses a spend-based tier system. New accounts start at Tier 1 (500 RPM, 30K TPM for GPT-4o). Tier 2 requires $50+ cumulative spend; Tier 3 requires $100+; Tier 4 requires $250+; Tier 5 requires $1,000+. Teams with sudden traffic spikes may hit rate limits before qualifying for the next tier — plan your usage ramp accordingly.

Tier Spend required GPT-4o RPM GPT-4o TPM
Tier 1 $0 500 30,000
Tier 2 $50 5,000 450,000
Tier 3 $100 5,000 800,000
Tier 4 $250 10,000 2,000,000
Tier 5 $1,000 10,000 30,000,000

Track your OpenAI spend automatically

PayMesh connects to your OpenAI account and syncs usage every hour — so you see actual costs before they hit your invoice. Set budget alerts and get notified before you overspend.

Compare providers

OpenAI is the widest-supported provider, but it's not always the cheapest or best-fit. Compare with:

See also: AI API Cost Comparison 2026 — total cost of ownership breakdown.