OpenAI API Pricing (March 2026) — GPT-5.4, O3 & GPT-4.1 Token Costs
OpenAI API pricing (March 2026): GPT-5.4 from $2.50/M, GPT-5.3 Codex at $3/$15, O3 Pro at $150. Compare GPT-5, GPT-4o, O1 pricing. Updated March 2026.
··3 min read
Share
OpenAI API Pricing — March 2026
Last updated: March 5, 2026 — GPT-5.4 now rolling out
Latest: GPT-5.4 (March 5, 2026)
- GPT-5.4 brings advances in reasoning, coding, and agentic workflows
- Available in ChatGPT, API, and Codex
- Includes GPT-5.4 Thinking and GPT-5.4 Pro variants
TL;DR — OpenAI API Prices (March 2026)
- GPT-5.2: $1.75/$14.00 per 1M input/output tokens (latest flagship)
- GPT-5: $1.25/$10.00 per 1M tokens
- GPT-5 Nano: $0.05/$0.40 per 1M tokens (cheapest)
- o4-mini: $1.10/$4.40 per 1M tokens (best value reasoning)
- o3-pro: $20.00/$80.00 per 1M tokens (strongest reasoning)
OpenAI offers the widest range of models from cheap fast models to expensive reasoning powerhouses.
GPT-5 Family (Flagship)
| Model | Context | Input / 1M tokens | Output / 1M tokens | Cached Input |
|---|---|---|---|---|
| GPT-5.2 Pro | 200K | $21.00 | $168.00 | $2.10 |
| GPT-5.2 | 200K | $1.75 | $14.00 | $0.175 |
| GPT-5.1 | 128K | $1.25 | $10.00 | $0.125 |
| GPT-5 | 128K | $1.25 | $10.00 | $0.125 |
| GPT-5 Mini | 200K | $0.25 | $2.00 | $0.025 |
| GPT-5 Nano | 128K | $0.05 | $0.40 | $0.005 |
O-Series (Reasoning)
| Model | Context | Input / 1M tokens | Output / 1M tokens | Cached Input |
|---|---|---|---|---|
| o1-pro | 200K | $150.00 | $600.00 | — |
| o3-pro | 200K | $20.00 | $80.00 | — |
| o1 | 200K | $15.00 | $60.00 | $7.50 |
| o3 | 200K | $2.00 | $8.00 | $1.00 |
| o4-mini | 200K | $1.10 | $4.40 | $0.275 |
| o3-mini | 200K | $1.10 | $4.40 | $0.275 |
GPT-4 Family (Previous Gen)
| Model | Context | Input / 1M tokens | Output / 1M tokens | Cached Input |
|---|---|---|---|---|
| GPT-4.1 | 1M | $2.00 | $8.00 | $0.20 |
| GPT-4.1 Mini | 1M | $0.40 | $1.60 | $0.04 |
| GPT-4.1 Nano | 1M | $0.10 | $0.40 | $0.01 |
| GPT-4o | 128K | $2.50 | $10.00 | $1.25 |
| GPT-4o Mini | 128K | $0.15 | $0.60 | $0.075 |
Cost Optimization Tips
- Use Nano/Mini models for simple tasks — 10-50x cheaper
- Prompt caching — 90% off for repeated context (cached = 10% of input price)
- Batch API — 50% off for async processing within 24 hours
- Choose the right model — Don't use o1-pro for simple queries
Related
Related Resources
ResourceLLM API Pricing (March 2026) — GPT-5.4, Claude, Gemini, DeepSeek & 30+ Models ComparedLLM API pricing trends 2026: GPT-5.4 from $2.50/M, Claude at $3/$15, DeepSeek at $0.14. Side-by-side cost comparison, optimization tips. Updated March 2026.ResourceClaude API Pricing (March 2026) — Opus 4.6, Sonnet 4.6, Haiku Token CostsUpdated March 2026. Anthropic Claude API pricing per 1M tokens: Opus 4.6 at $5/$25, Sonnet 4.6 at $3/$15, Haiku at $0.25/$1.25. Full cost table and optimization tips.ResourceDeepSeek API Pricing (March 2026) — V3.2 & R1 Reasoner CostsDeepSeek API pricing (March 2026): V3.2 at $0.28/$0.42 per 1M tokens, R1 at $0.50/$2.18. Cheapest LLM API. Full cost table, free tier, and comparison with GPT and Claude.ResourceGemini API Pricing (March 2026) — 2.5 Pro, Flash & Free Tier Token CostsUpdated March 2026. Google Gemini API pricing per 1M tokens: 2.5 Pro at $1.25/$10, Flash at $0.30/$2.50, Flash-Lite at $0.10/$0.40. Free tier on most models.
Want more resources?
Subscribe to get the latest AI tools, guides, and updates.
Newsletter
Stay ahead of the curve
Key insights from top tech podcasts, delivered daily. Join 10,000+ engineers, founders, and investors.
One email per day. Unsubscribe anytime.