
1 / 3
【Overview】 LLMBoard is an independent leaderboard and directory for comparing AI models. It covers language, image, video, speech and embedding models, bringing together benchmark capability, official API pricing, runtime performance and context size in one place. The goal is to help users cut through marketing claims and see how models actually stack up on the dimensions that matter for their work. The directory catalogs 1,346 models with input/output modality, price, ranking status and release dates, filterable by vendor, modality and price. Alongside it, an AI benchmark directory lists 806 benchmarks by category, family, metric, coverage and evidence grade, with per-benchmark pages showing model results. This makes LLMBoard useful for anyone who needs to shortlist models for a specific task — such as coding, reasoning, math or embeddings — rather than relying on a single headline score. LLMBoard is free to use. It is aimed at developers, technical teams and researchers who want a structured, evidence-based view of the AI model landscape before committing to an API or building a workflow around a particular model. 【Key features】 • Overall capability leaderboard: Ranks models by an aggregated capability score built from benchmark evidence, coverage, official pricing and availability. • Task-specific leaderboards: Separate rankings for coding, reasoning, math, knowledge, instruction following, text, vision, agents, open models and embeddings. • LLM pricing leaderboard: Ranks models by official vendor pay-as-you-go API price in USD per one million tokens. • Modality pricing pages: Break out pricing comparisons for image, video and audio models in addition to chat token pricing. • Runtime and latency data: Publishes output speed, catalog latency, observed TTFT, throughput and provider reliability in a standalone runtime view. • AI benchmark directory: Lists 806 benchmarks by category, family, metric, coverage and evidence grade, with per-benchmark pages of model results. • Model directory: Catalogs 1,346 language, image, video, speech and embedding models with input/output modality, price, ranking status and release dates, filterable by vendor, modality and price. • Scoring methodology documentation: Explains how scores are aggregated, how missing results and related benchmarks are handled, how model aliases are matched, and what evidence coverage means. 【Use cases】 1. Find the best LLM for coding. 2. Find the best model for reasoning, math or knowledge tasks. 3. Compare official API token prices across vendors. 4. Compare image, video and audio model pricing. 5. Check model context window sizes. 【Pros and cons】 Pros: - The independent stance and documented methodology make it a credible first stop for shortlisting models, especially for teams that need to justify a choice to others. - Separate task-specific rankings mean you are not stuck with a single general score when your priority is coding, math or embeddings. Cons: - Capability scores are for screening, not a verdict: a score of 97 is not seven percent more capable than a score of 90. - Pricing data reflects only the model vendor's standard pay-as-you-go API rate; third-party offers are not substituted, and unavailable prices show as N/A. - Runtime figures are reported separately and are not folded into capability scores. - Benchmark and LM Arena values retain their original scales rather than being normalized. - The model directory includes products without a comparative LLMBoard score, so catalog presence alone does not indicate a ranking. 【FAQ】 Q: Can I compare API prices for image, video and audio models on LLMBoard? A: Yes. In addition to chat token pricing, LLMBoard has modality pricing pages that break out pricing comparisons for image, video and audio models. Q: Does LLMBoard factor speed and latency into its capability scores? A: No. Output speed, catalog latency, observed TTFT, throughput and provider reliability are published in a standalone runtime view and are not folded into capability scores. Q: How does LLMBoard handle models that are in the directory but not ranked? A: The model directory includes products without a comparative LLMBoard score — these are catalog-only entries. The directory also lists models with ranking status, so you can distinguish ranked from unranked models.
free
WPS AI is an AI assistant built into WPS Office that helps users write documents, generate presentation outlines and…
Baidu Comate is an AI coding assistant built by Baidu on its ERNIE model. It goes beyond simple autocomplete: it…
UniConverter is an AI media conversion and enhancement tool for video, audio, and images. It combines everyday…
Added
10/20/2026
Views
851
Clicks
4
6 similar tools to consider instead of LLMBoard