AI API Pricing May 2026
This May 2026 refresh updates LangCopilot's pricing catalog around the models teams are most likely to search for now: OpenAI GPT-5.5, Anthropic Claude Opus 4.7, Google Gemini 3.1, xAI Grok 4.3 and Grok 4.20, DeepSeek V4, Qwen3.6 Plus, and Kimi K2.6.
The main change is not just "more models." The pricing surface is now more complex: long-context thresholds, batch/flex/priority modes, cache-hit rates, and temporary discounts can change the winner for the same workload.
Quick price table
| Model | Status | Context | Input / 1M | Cached input / 1M | Output / 1M |
|---|---|---|---|---|---|
| GPT-5.5 | active | 1M | $5.00 | $0.50 | $30.00 |
| Claude Opus 4.7 | active | 1M | $5.00 | $0.50 | $25.00 |
| Gemini 3.1 Pro Preview | preview | 1M | 4.00 >200K | 0.40 >200K | 18.00 >200K |
| Grok 4.3 | active | 1M | $1.25 | $0.20 | $2.50 |
| DeepSeek V4 Pro | active | 1M | $0.435 | $0.003625 | $0.87 |
| Qwen3.6 Plus | active | 1M | 2.00 >256K | not listed | 6.00 >256K |
| Kimi K2.6 | active | 256K | $0.95 | $0.16 | $4.00 |
Official sources: OpenAI pricing, Anthropic pricing, Gemini API pricing, xAI pricing, DeepSeek pricing, Qwen Cloud pricing, and Kimi API Platform.
What changed
- OpenAI GPT-5.5 and GPT-5.4 pricing now needs short-context, long-context, batch, flex, and priority treatment.
- Anthropic Claude Opus 4.7 adds a new premium Anthropic route with 1M context and optional fast mode.
- Gemini 3.1 Pro Preview has a clear 200K prompt threshold, so calculator math must switch rates above that point.
- xAI's current search intent should move toward Grok 4.3 and Grok 4.20 instead of old Grok 4 pages.
- DeepSeek V4 Pro is temporarily discounted through 2026-05-31 15:59 UTC, making it a strong cost comparison candidate.
- Qwen3.6 Plus requires tiered prompt pricing rather than one flat input/output rate.
Recommended calculator routes
Start from the pricing hub, then open the model-specific calculators:
- GPT-5.5 token calculator
- Claude Opus 4.7 token calculator
- Gemini 3.1 Pro token calculator
- Grok 4.3 token calculator
- DeepSeek V4 Pro token calculator
- Qwen3.6 Plus token calculator
- Kimi K2.6 token calculator
SEO update note
The safest SEO move is to stop indexing every possible pairwise comparison. Keep the high-intent compare pages, noindex retired model routes, and connect the new pricing data to editorial pages like this one. That gives crawlers a clearer reason to index the pages that matter.