limited
Anthropic's limited-availability Claude variant for approved Project Glasswing customers.
- Context
- 1M tokens
- Weights
- closed
- Input / 1M
- US$10.00
- Output / 1M
- US$50.00
Source checked4 sourcesranked
Model directory
Browse exact source-qualified variants by provider, model type, reasoning mode and availability. Cards link to complete evidence profiles and comparisons.
103 variants · checked 15 July 2026 · 102 ranked
Anthropic's limited-availability Claude variant for approved Project Glasswing customers.
Anthropic's most capable widely released model, documented for long-running agent workloads.
Moonshot AI's 2.8T-parameter frontier model with native vision, a 1M-token context window, and API pricing from the Kimi K3 launch; open weights scheduled after launch.
OpenAI's documented flagship API model for complex reasoning and coding workloads.
Anthropic's documented model for complex agentic coding and enterprise work.
xAI's documented chat API model for coding, agentic tasks and knowledge work.
Meta Muse Spark 1.1 frontier API model with configurable reasoning, 1M context, and strong agentic scores on provider and independent evaluations.
OpenAI's documented frontier API model with configurable reasoning and a 1.05-million-token context window.
OpenAI's documented GPT-5.6 model positioned between frontier capability and API cost.
Anthropic's Claude Opus 4.7 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Z.AI's GLM-5.2 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Anthropic's current Sonnet model, documented as a balance of speed and intelligence.
Alibaba Cloud's documented flagship Qwen model for complex reasoning and agent workloads.
Alibaba Cloud's documented Qwen model balancing capability, cost and built-in tools.
OpenAI's documented reasoning model for coding and complex professional work.
Google's Gemini 3 Pro model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Google's stable multimodal model for agentic, coding and high-throughput workloads.
Xiaomi's MiMo-V2-Pro model tracked via the BenchLM July 2026 leaderboard for capability comparison.
OpenAI's GPT-5.2 Pro model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Xiaomi's MiMo-V2.5-Pro model tracked via the BenchLM July 2026 leaderboard for capability comparison.
OpenAI's documented GPT-5.6 model for cost-sensitive, high-volume API workloads.
OpenAI's API model optimized for agentic coding tasks in Codex and similar environments.
MiniMax's current long-context multimodal API model for coding and agentic workloads.
Anthropic's Claude Sonnet 4.6 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Anthropic's Claude Opus 4.6 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
OpenAI's GPT-5.5 Pro model tracked via the BenchLM July 2026 leaderboard for capability comparison.
GLM-4.7 is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Z.AI's GLM-5-Turbo model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Z.AI's GLM-5.1 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Anthropic's Claude Opus 4.5 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Google's Gemini 3 Pro Deep Think model tracked via the BenchLM July 2026 leaderboard for capability comparison.
OpenAI's GPT-5.1 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
MiniMax's MiniMax M2.7 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
xAI's Grok 4.1 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Z.AI's GLM-5V-Turbo model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Kimi K2.6 is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Alibaba's Qwen3.6 Plus model tracked via the BenchLM July 2026 leaderboard for capability comparison.
GPT-5.2 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Z.AI's GLM-5 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
xAI's documented general chat model with configurable reasoning and a one-million-token context window.
Xiaomi's MiMo-V2-Omni model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Google's preview model for advanced multimodal understanding and agentic workloads.
Meta Muse Spark predecessor to Muse Spark 1.1 with a 262K context window; list API pricing not published on checked sources.
Kimi K2.7 Code is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Grok 4 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
MiMo-V2.5 is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
DeepSeek V3.2 is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
MiMo-V2-Flash is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Claude Sonnet 4.5 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
DeepSeek's official V4 API model with optional thinking, tool use and open model weights.
Qwen3.5 397B is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
GPT-5.4 mini is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Moonshot AI's officially documented multimodal Kimi model; unverified K3 reports are excluded from this registry.
OpenAI's GPT-5.4 nano model tracked via the BenchLM July 2026 leaderboard for capability comparison.
GLM-4.6 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Google's Gemini 3 Flash model tracked via the BenchLM July 2026 leaderboard for capability comparison.
DeepSeek's official lower-cost V4 API model with optional thinking and open model weights.
MiniMax M2.5 is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Gemini 2.5 Pro is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Grok 4.20 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Google's Gemma 4 31B model tracked via the BenchLM July 2026 leaderboard for capability comparison.
o3 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
GPT-5 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
o3-pro is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Tencent Hy Team open-weight MoE (295B total / 21B active) with hybrid reasoning, 256K context, and Apache 2.0 weights released July 2026.
Anthropic's documented low-latency Claude model with text and image input.
GPT-5 mini is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Google's stable low-latency multimodal model for high-frequency lightweight tasks.
Qwen3.6 27B is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Nemotron 3 Ultra is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Gemini 2.5 Flash is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Qwen3 Max is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
GPT-5 Codex is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Command A+ is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
o1 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Qwen3.6 Max is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
GPT-5.2 Codex is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Qwen3.5 122B is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Claude Opus 4.1 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
DeepSeek V3.1 Terminus is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Claude Opus 4 is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Mistral AI's open hybrid model combining instruction following, reasoning and coding.
Gemma 4 12B is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Claude Sonnet 4 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Kimi K2 Thinking is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Mistral AI's open-weight multimodal model for agentic and coding use cases.
Mistral AI's documented open-weight general-purpose multimodal model.
Grok Code Fast 1 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Nova 2 Pro is a API model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
MiniMax M2.1 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Qwen3.5 27B is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Grok 4 Fast is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
MiniMax M2 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Gemma 4 26B is a open-weight model tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, and reasoning.
Claude Sonnet 3.7 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
o4-mini is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Nemotron 3 Super is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Qwen3.5 Omni Plus is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Qwen3.5 35B is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Ling 2.6 1T is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Grok 3 mini is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Kimi K2 0905 is tracked via Artificial Analysis independent evaluations for capability comparison across coding, agents, reasoning and related workloads.
Alibaba's July 2026 preview of its 2.4T-parameter flagship Qwen3.8 Max line. Identity and preview availability are source-checked; no public benchmark table was published at check date, so ranking evidence remains not evaluated.