Frontier Benchmarks AI · Changelog https://frontierbenchmarks.com/changelog Changes in the frontier LLM atlas — data updates, new features, methodology changes. en Fri, 01 May 2026 00:00:00 GMT [product] Cross-provider Pricing Calculator https://frontierbenchmarks.com/pricing https://frontierbenchmarks.com/changelog#2026-05-01-0 Fri, 01 May 2026 00:00:00 GMT product Estimated monthly TCO per model + provider. Inputs: M tokens input/output/cached. Compares up to 12 models sorted ascending by cost. Cache support when the provider exposes it. [product] Use Case Wizard recommender https://frontierbenchmarks.com/wizard https://frontierbenchmarks.com/changelog#2026-05-01-1 Fri, 01 May 2026 00:00:00 GMT product 4-step wizard to recommend the top 3 models per use case (coding, math, writing, vision, agent, RAG, summarization, translation). Composite = baseScore × coverage × priorityFactor. [product] Head-to-head Battle Mode https://frontierbenchmarks.com/battle https://frontierbenchmarks.com/changelog#2026-05-01-2 Fri, 01 May 2026 00:00:00 GMT product Side-by-side comparison of 2-4 models benchmark by benchmark. Shareable URL with ?models=a,b,c. Global verdict with winner by wins, abstentions when no score. [product] Hardware Compatibility Checker https://frontierbenchmarks.com/hardware https://frontierbenchmarks.com/changelog#2026-05-01-3 Fri, 01 May 2026 00:00:00 GMT product Detects browser GPU/RAM/CPU (with honest Firefox/Safari limitations) and classifies each model into S/A/B/C/D/F tiers based on available VRAM and quantization. Multi-GPU 1-8. [data] Open-source model enrichment https://frontierbenchmarks.com/changelog https://frontierbenchmarks.com/changelog#2026-04-30-4 Thu, 30 Apr 2026 00:00:00 GMT data Added total params, active params (MoE), license, cross-provider pricing and install commands (Ollama / LM Studio / vLLM) to the main open-weight models. [methodology] Definition of "comparable" in battles https://frontierbenchmarks.com/methodology https://frontierbenchmarks.com/changelog#2026-04-29-5 Wed, 29 Apr 2026 00:00:00 GMT methodology A benchmark is comparable only if 2+ models in the battle have a published score. If a model has no score, it counts as abstained (does not affect winRate). [product] Frontier Benchmarks AI launch https://frontierbenchmarks.com/models https://frontierbenchmarks.com/changelog#2026-04-28-6 Tue, 28 Apr 2026 00:00:00 GMT product First public version of the atlas. 62 models, 32 benchmarks, 25 companies. Catalog, individual Models, Benchmarks, Companies, Methodology and Download single-file views.