Top 10 Cheapest AI APIs in 2026 (Ranked by Real Cost)
Independent cost ranking of 10 major LLMs in 2026. Per-call price comparison, where each model wins, and how caching can change the order entirely.
"Cheapest AI API" is a misleading question. The model that costs the least per token might be useless for your task — and the one that looks expensive might be 10× cheaper for what you actually use it for. So before we hand you the list, two caveats:
- Cost is meaningless without capability matching. A $0.20/1M model that gets 60% of your queries wrong is more expensive than a $5/1M model that nails them on the first try.
- Headline rates lie in 2026. Caching can cut bills by 90%. Batch API drops them 50%. The "cheapest" model on the price page might be the most expensive in production.
With those out of the way: here's the honest ranking by single-call cost (1,000 input + 500 output tokens) across the 10 frontier and small models on this site.
Methodology
Each cost figure is calculated as:
cost = (1,000 / 1,000,000) × input_price + (500 / 1,000,000) × output_price
Where input_price and output_price are the official 2026 published rates per 1M tokens. The numbers don't include caching or batch discounts — those are footnoted because they change the order substantially.
All 10 models are pre-loaded into the calculator on the homepage so you can plug in your own ratios.
The Ranking
| Rank | Model | Provider | Per-call cost | Best for |
|---|---|---|---|---|
| 1 | GPT-5 mini | OpenAI | $0.0006 | Default everyday small |
| 2 | Gemini 3.0 Flash | $0.0013 | Multimodal at scale | |
| 3 | DeepSeek V4 | DeepSeek | $0.0009 | Coding, math, reasoning value |
| 4 | o4-mini | OpenAI | $0.0027 | STEM reasoning |
| 5 | Claude Haiku 4.5 | Anthropic | $0.0035 | Anthropic ecosystem, caching-heavy workloads |
| 6 | Gemini 3.0 Pro | $0.0075 | Long context (2M tokens) | |
| 7 | Mistral Large 3 | Mistral | $0.0058 | EU hosting, multilingual |
| 8 | Grok 4 | xAI | $0.0140 | Real-time X integration |
| 9 | GPT-5.5 | OpenAI | $0.0150 | Frontier multimodal |
| 10 | Claude Opus 4.7 | Anthropic | $0.0525 | Hard reasoning, 1M context |
Numbers update monthly. The figures above reflect rates verified May 2026. The calculator always shows the current snapshot.
#1: GPT-5 mini ($0.0006/call)
OpenAI's small model is the new default for high-volume production. At $0.20 input / $0.80 output per 1M tokens, it's:
- 25× cheaper than GPT-5.5
- 60% cheaper than Haiku 4.5
- 30% cheaper than Gemini 3.0 Flash on output
Where it wins: chatbots, classification, function calling, vision tasks at moderate complexity. With prompt caching (cached input at $0.05/1M), volume workloads get even cheaper.
Where it loses: hard reasoning (use o4-mini instead), long context (use Gemini 3.0 Pro).
See the full GPT-5 mini calculator →
#2: DeepSeek V4 ($0.0009/call)
The most aggressive cost/quality story in 2026. DeepSeek V4 is an open-weight 1T-parameter MoE that punches at the level of US frontier models on coding and reasoning at 3% of GPT-5.5's price.
Trade-offs:
- China-based; some enterprises have data residency concerns
- Slightly weaker on creative writing and English nuance
- No vision (yet)
If you're cost-sensitive and your workload is coding, math, or reasoning-heavy, DeepSeek V4 is the rational pick. The savings vs comparable Western models can fund 10× the volume.
See the full DeepSeek V4 calculator →
#3: Gemini 3.0 Flash ($0.0013/call)
Google's high-throughput multimodal model. The pitch:
- Native audio + vision (no separate model needed)
- 1M token context window
- Fast inference (multi-thousand tokens/sec)
- Caching support
For multimodal pipelines (image classification, audio summarization, document QA), Gemini 3.0 Flash is the sweet spot. For pure text, GPT-5 mini is slightly cheaper but less capable on long context.
See the full Gemini 3.0 Flash calculator →
#4: o4-mini ($0.0027/call)
OpenAI's reasoning model — descended from o1 / o3. At $0.90 input / $3.60 output, it's 5× more expensive than GPT-5 mini but punches multiple weight classes above on:
- STEM problems (math, physics, chemistry)
- Multi-step coding refactors
- Logic puzzles requiring chain of thought
Use when: the task genuinely needs reasoning. Don't use for chatbots — you'll burn 5× the cost for no quality gain.
See the full o4-mini calculator →
#5: Claude Haiku 4.5 ($0.0035/call)
Anthropic's small model is 3× more expensive than GPT-5 mini at face value — but with caching, the math inverts.
Haiku's cached input price is $0.10/1M (vs GPT-5 mini's $0.05). Both cheap. But Haiku's relative discount vs its standard input ($1.00) is 90% off — meaning for cache-heavy workloads, Haiku 4.5 becomes one of the cheapest models in the entire lineup.
The classic example: a chatbot with a 2,000-token system prompt called millions of times. With 95% cache hit rate:
- Standard cost: $1.90 per 1,000 calls
- With caching: ~$0.30 per 1,000 calls
Claude Haiku 4.5 also has stronger reasoning per dollar than GPT-5 mini on nuanced tasks. If you're already in the Anthropic ecosystem, it's hard to beat.
See the full Claude Haiku 4.5 calculator →
#6-#7: Mid-tier flagships
Mistral Large 3 ($0.0058/call) and Gemini 3.0 Pro ($0.0075/call) sit in an awkward middle: more expensive than the small models but considerably cheaper than the absolute frontier.
Mistral Large 3: Best for EU customers with data residency requirements. Multilingual is its strongest pitch — handles 30+ European languages natively.
Gemini 3.0 Pro: The 2M token context is unmatched. If you're doing book-length analysis, entire codebase review, or long video understanding, it's the only practical option.
#8-#9: Premium flagships
Grok 4 ($0.0140/call) is the wildcard. Its real-time X (Twitter) integration is unique — you can ask it about what's happening right now in tech, sports, finance. Premium price reflects this niche feature; for general tasks it's hard to justify over GPT-5.5.
GPT-5.5 ($0.0150/call) is the all-rounder frontier model. When in doubt, this is the safe choice — best ecosystem support, best tooling, best documentation, broad capability. Premium pricing reflects the ecosystem moat as much as the model itself.
#10: Claude Opus 4.7 ($0.0525/call)
The most expensive model on this list — by a significant margin. 3.5× more expensive per call than GPT-5.5.
So why use it? Three reasons:
- Hard reasoning: Claude Opus consistently leads on multi-step coding, agentic workflows, and complex analysis.
- 1M token context with cleaner long-context attention than the alternatives.
- Caching changes everything: Opus 4.7's cached read price is $1.50/1M — the same as GPT-5.5's standard input. With heavy caching, Opus's effective cost drops dramatically.
For a code review agent calling Opus 1,000×/day with a 5K-token cached system prompt, the caching savings can hit 70-80%. The "expensive" model becomes competitive with the small models — but at frontier-tier quality.
See the full Claude Opus 4.7 calculator →
What changes the order?
The ranking above is for naive single-call cost. Three things substantially change which model is actually cheapest for your use case:
1. Caching ratio
If 80% of your input is cached (a typical RAG application), the order shifts dramatically:
| Model | Naive cost | With 80% caching | Order shift |
|---|---|---|---|
| GPT-5 mini | $0.0006 | $0.00048 | unchanged |
| Claude Haiku 4.5 | $0.0035 | $0.00094 | jumps from #5 to #2 |
| Claude Opus 4.7 | $0.0525 | $0.0156 | jumps from #10 to #5 |
2. Output ratio
If you're generating long content (output >> input), output prices dominate. Models with cheap output (Gemini 3.0 Flash at $2/1M, GPT-5 mini at $0.80/1M) become disproportionately cheaper.
3. Batch eligibility
If your workload tolerates 24-hour async processing, Batch API discounts cut all of OpenAI / Anthropic / Google rates by 50%. Models without Batch support (DeepSeek, Grok 4) lose this lever.
How to actually pick a model
A practical decision tree:
- Is the task complex reasoning? → o4-mini for cost, Opus 4.7 for quality
- Is the context > 200K tokens? → Gemini 3.0 Pro
- Is the workload cache-heavy with stable prompts? → Haiku 4.5 (best cache discount)
- Is the workload batchable (non-realtime)? → Anything with batch + 50% off
- Default high-volume simple tasks? → GPT-5 mini or Gemini 3.0 Flash
- Need EU hosting? → Mistral Large 3
- Cost is the only concern, quality acceptable? → DeepSeek V4
Calculate your real cost
The ranking above assumes 1,000 input + 500 output tokens. Your workload is different.
Use the calculator on the homepage → — pick your model, plug in your token counts, toggle caching and batch options. The "cheapest model for your specific case" appears at the top of the comparison table.
If you're spending more than $500/month on AI APIs and haven't run this exercise, you're almost certainly leaving 30-60% on the table.
Pricing reflects rates verified May 2026. Verify with each provider's official pricing page before committing budget. The models.json on GitHub shows last verification dates.