Pricing
The homepage does not list subscription pricing for using the directory itself; the only disclosed pricing is for sponsor placements, which start at $149 per month for a homepage logo slot.
Learn moreLMSpeed
Compare LLM API pricing, speed and security across 100+ providers before you commit.
Intermediate
What is LMSpeed?
LMSpeed is a free directory and benchmark hub that tracks LLM model pricing, throughput, latency and Artificial Analysis scores across 100+ API providers, plus leaderboards and security audit reports for choosing a provider.
LMSpeed is a model and API provider directory built for anyone deciding which large language model or API relay to use for a project. It tracks newly released models alongside established ones, recording context window size, input and output price per million tokens, number of providers offering each model, throughput, latency and release date. Scores from Artificial Analysis cover overall intelligence, coding and math, giving a quick read on how a model performs relative to peers before any code is written. The site organizes this data into leaderboards for specific jobs: agents, coding and reasoning, plus monthly rankings for latency, throughput, provider health and pricing. A separate audit section reviews third-party API relays and scores each one on endpoint profile, model identity, prompt safety and response integrity, flagging providers where prompts, model identity or responses may leak or degrade. Individual audit reports show a report date and a breakdown score out of 100 for a specific endpoint and model pairing. The tool suits developers, technical buyers and teams evaluating which provider to route traffic through, especially when the same model is resold by many relays at different prices and reliability levels. It works best as a research step before signing a contract or wiring up an SDK, rather than as a runtime monitoring tool. Compared to a single provider's own pricing page, LMSpeed's value is in laying multiple providers for the same model side by side, surfacing cheaper or faster alternatives that might otherwise be missed, and adding a security lens that most pricing comparison sites skip.
How do you use LMSpeed?
1Open the model directory
Go to the model directory to see every tracked model with context window, token pricing, provider count, throughput and latency in one table.
2Filter for newest or strongest releases
Switch between the newest releases and the strongest models view to spot newly tracked models and the current Artificial Analysis leaders for intelligence, coding and math.3Compare a model's pricing and providers
Open an individual model page to see its input and output price per million tokens, the list of providers offering it, and its release date and license.
4Check license terms
Review the license page linked from each model to confirm whether it is closed proprietary or open weights before committing to a provider.
5Review API provider security audits
Browse the audit report list to see endpoint profile, model identity, prompt safety and response integrity scores for specific provider relays before signing up.
6Inspect a single audit report
Open an individual audit report to see the scored breakdown for a specific endpoint and model pairing, including report date and total score.
Pros and cons
Pros
- Free access to a large model directory covering pricing and performance data from 100+ providersAI
- Combines pricing, throughput, latency and benchmark scores in a single comparison tableAI
- Dedicated security audit reports flag prompt, model identity and response integrity risks before committing to a providerAI
- Multiple leaderboards for agents, coding, reasoning, latency, throughput and pricing narrow choices for specific use casesAI
- Tracks newly released models quickly, useful for staying current on fast moving releasesAI
Cons
- Many benchmark scores carry wide error margins, so rankings can shift noticeably between updatesAI
- Audit reports only cover a sample of third party relay endpoints, not every provider a buyer might considerAI
- No visible documentation of how often underlying data is refreshed or independently verifiedAI
- Homepage does not disclose any account tiers or pricing for the platform itself beyond sponsor placementsAI
- Dense tables of metrics may overwhelm users who are new to LLM benchmarking terminologyAI
How much does LMSpeed cost?
Features
A model directory listing context window, input and output price per million tokens, provider count, throughput, latency, release date and Artificial Analysis scores for intelligence, coding and math; leaderboards for agents, coding, reasoning, latency, throughput, pricing and provider health; and audit reports scoring API relay security and response integrity.
Learn moreFrequently asked questions about LMSpeed
How much does LMSpeed cost?
The homepage does not list subscription pricing for using the directory itself; the only disclosed pricing is for sponsor placements, which start at $149 per month for a homepage logo slot.How do you use LMSpeed?
The walkthrough on this page covers 6 steps: 1. Open the model directory 2. Filter for newest or strongest releases 3. Compare a model's pricing and providers 4. Check license terms 5. Review API provider security audits 6. Inspect a single audit report.What platforms does LMSpeed support?
LMSpeed is available on Web App.What are the limitations of LMSpeed?
Many benchmark scores carry wide error margins, so rankings can shift noticeably between updates. Audit reports only cover a sample of third party relay endpoints, not every provider a buyer might consider. No visible documentation of how often underlying data is refreshed or independently verified.
