As of August 30, 2026, LFM2.5-2.6B is 99% cheaper per million tokens than Claude 3.5 Haiku. Claude 3.5 Haiku is $0.80 / $4.00 per million input/output tokens versus LFM2.5-2.6B at $0.02 / $0.02. Sources: https://platform.claude.com/docs/en/about-claude/pricing and https://www.liquid.ai
Sources: Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Liquid AI pricing (https://www.liquid.ai)
Claude 3.5 Haiku
Anthropic
Input
$0.80
Output
$4.00
LFM2.5-2.6B
Liquid AI
Input
$0.02
Output
$0.02
Token pricing, specs, and benchmarks side by side
Each bar is share of the current leader on that eval. Click a row to open the board.
AIME 2025. Left is cheaper. Up is a higher score. The line is the best score you can buy at each price.
Claude 3.5 HaikuLFM2.5-2.6BVerdict
LFM2.5-2.6B offers both lower pricing and stronger benchmark performance across the board, making it the clear value leader in this comparison.
As of August 30, 2026, LFM2.5-2.6B is 99% cheaper on blended LLM API pricing. LFM2.5-2.6B costs $0.02 per million input tokens and $0.02 per million output tokens ($0.04 blended 1M-in + 1M-out). Claude 3.5 Haiku costs $0.80 / $4.00 per million tokens ($4.80 blended). Sources: https://platform.claude.com/docs/en/about-claude/pricing and https://www.liquid.ai. The cheapest LLM for your app still depends on how many output tokens you generate.
Claude 3.5 Haiku Claude API pricing from Anthropic is $0.80 input and $4.00 output per million tokens via Anthropic. LFM2.5-2.6B Liquid AI API pricing is $0.02 input and $0.02 output per million tokens via Liquid AI. Use those four numbers, not ChatGPT Plus or Claude Pro subscription prices, when you are comparing APIs.
Claude 3.5 Haiku has a published SWE-bench Verified score of 40.6%. LFM2.5-2.6B does not have that eval on this page yet.
Claude 3.5 Haiku supports a 200K token context window, while LFM2.5-2.6B supports 66K tokens. Claude 3.5 Haiku offers a larger context window. Filling a larger window bills more input tokens, so the cheaper-per-million model can still cost more on long documents.
LFM2.5-2.6B leads on 1 of 1 shared benchmarks versus Claude 3.5 Haiku's 0 wins. Check the comparison table for GPQA Diamond, SWE-bench, MMLU, HLE, and the other evals we track. Token price and benchmark score together are the usual LLM comparison, not either number alone.
Both Claude 3.5 Haiku and LFM2.5-2.6B are production API models. For cost-sensitive production traffic, LFM2.5-2.6B has better economics on the blended 1M-in + 1M-out scale. For maximum capability, weight the benchmark table for your domain. Many production stacks route cheap models for drafts and a frontier model for the hard turn, which is usually cheaper than sending everything to the expensive API.
Yes. Call Claude 3.5 Haiku through Anthropic and LFM2.5-2.6B through Liquid AI with separate API keys, or through a gateway that already wraps both. Keep prompts in tokens, not characters, when you estimate the invoice. AnotherWrapper templates can swap providers without rewriting auth and billing.
List rates are the latest published Claude API pricing from Anthropic and Liquid AI API pricing, quoted per million tokens. Official rate sheets: https://platform.claude.com/docs/en/about-claude/pricing and https://www.liquid.ai. Prompt caching, batch APIs, and committed-use discounts can change the invoice. Benchmark scores are from official publications and independent evals, updated as new numbers land.
Claude 3.5 Haiku supports up to 200K output tokens per request, while LFM2.5-2.6B supports up to an unspecified number of output tokens. Output tokens are usually the expensive half of API pricing, so a higher max output cap is a capability, not a discount.
Claude 3.5 Haiku has a measured throughput of approximately 104 tokens/second. Throughput data for LFM2.5-2.6B is not yet available. Throughput (tokens per second) and time-to-first-token are why two models with the same per-million-token price can still feel different in chat and agents.
Use the model pickers at the top of this LLM comparison. Search any model in the index and the URL updates. Open Claude 3.5 Haiku or LFM2.5-2.6B from the sidebar for the single-model API pricing page, or go back to the LLM pricing table to sort by cheapest blended cost.
This comparison covers API pricing, context window, throughput, and benchmark scores. Claude 3.5 Haiku costs $0.80/M input and $4.00/M output. LFM2.5-2.6B costs $0.02/M input and $0.02/M output.
Rank both on the LLM leaderboard, SWE-bench, and Humanity's Last Exam.
The index
343 models across 36 providers. Search or jump to a lab — every model page stays linked here.
Next step
Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.
