AI · LEADERBOARD

The AI Leaderboard

Q: What is the best AI model right now?

As of 14 July 2026, Claude Fable 5 (with fallback) by Anthropic is the highest-ranked model on our daily intelligence index. The leaderboard re-ranks every day, so check back for the current pick.

Q: What is the best open-source AI model?

GLM-5.2 (max) is currently the strongest open-source model in our ranking, the top pick for teams that want to self-host or keep data sovereign.

Q: Which AI model has the best price-to-performance?

MiMo-V2.5 currently scores highest on our value metric, which weighs intelligence and speed against the cost per million tokens.

Q: Which AI model has the largest context window?

Muse Spark 1.1 (xhigh) leads on context window at 1M tokens, useful for long documents, large codebases and extended conversations.

Q: How often is the leaderboard updated?

Every day at 04:00 Europe/Amsterdam. We pull fresh public benchmark data and re-rank, with no smoothing and no manual edits.

Q: How are the AI models ranked?

We aggregate public benchmarks into an intelligence index, then score value as intelligence weighed against output speed and cost per million tokens. Output cost is weighted more heavily because real usage is output-heavy.

Q: Is the SevenLab AI leaderboard free and independent?

Yes. The ranking is free, published under Creative Commons, and fully independent. We take no vendor payments to place models, these are the picks our own teams ship with.

Daily ranking of the strongest AI models. No vendor spin, no marketing pages, just the picks teams actually ship with.

As of 14 July 2026, the top-ranked AI model is Claude Fable 5 (with fallback) by Anthropic. The strongest open-source model is GLM-5.2 (max), and MiMo-V2.5 offers the best price-to-performance. Rankings refresh every day at 04:00 Europe/Amsterdam.

Best closed model

Claude Fable 5 (with fallback)

Anthropic

Intel: 59.9
Speed: 62t/s
Cost in/out: €8.75 / €43.74

State-of-the-art reasoning and complex agentic workflows where budget is no barrier.

Best open model

GLM-5.2 (max)

Z AI

Intel: 51.1
Speed: 208t/s
Cost in/out: Provider€1.22 / €3.85Sovereign~€66 / hr*

Top-tier open-source deployment for sovereign enterprise applications requiring maximum intelligence.

Best value177.2

MiMo-V2.5

Xiaomi

Intel: 37.2
Speed: 88t/s
Cost in/out: Provider€0.12 / €0.24Sovereign~€33 / hr*

Extremely cheap structured data extraction with a large context window.

Top 10, side by side

#↑	Model	Vendor	Intelligence	Speed (tok/s)	Cost (€/Mtok)	Context	Best for
1	Claude Fable 5 (with fallback)	Anthropic	59.9	62	€8.75 / €43.74	1.0M	State-of-the-art reasoning and complex agentic workflows where budget is no barrier.
2	GPT-5.6 Sol (max)	OpenAI	58.9	76	€4.37 / €26.25	1.0M	Ultra-high intelligence tasks requiring massive reasoning capabilities and OpenAI ecosystem integration.
3	Claude Opus 4.8 (max)	Anthropic	55.7	55	€4.37 / €21.87	1.0M	Deep academic research and highly nuanced text analysis at a premium price point.
4	GPT-5.6 Terra (max)	OpenAI	55.0	155	€2.19 / €13.12	1.0M	High-speed, high-intelligence enterprise automation requiring rapid throughput.
5	GPT-5.5 (xhigh)	OpenAI	54.8	81	€4.37 / €26.25	922K	Complex reasoning and multi-step problem solving with a large context window.
6	Grok 4.5 (high)	SpaceXAI	53.8	114	€1.75 / €5.25	500K	Real-time information processing and high-speed reasoning at a competitive price.
7	Claude Opus 4.7 (max)	Anthropic	53.5	52	€4.37 / €21.87	1.0M	Nuanced literary analysis and complex document synthesis with high reliability.
8	Claude Sonnet 5 (max)	Anthropic	53.4	79	€1.75 / €8.75	1.0M	Balanced enterprise workflows requiring high intelligence and moderate operational costs.
9	GPT-5.6 Luna (max)	OpenAI	51.2	234	€0.87 / €5.25	1.0M	High-volume, low-latency customer support and real-time conversational agents.
10	ZGLM-5.2 (max)	Z AI	51.1	208	€1.22 / €3.85 ~€66 / hr*	1.0M	Top-tier open-source deployment for sovereign enterprise applications requiring maximum intelligence.

Sovereign deploy cost is hourly (Scaleway dedicated GPU cluster, EU, all-in SevenLab pricing). Hover any estimate to see the exact GPU profile. Final price depends on throughput and infra choices.

#1 · proprietary

Claude Fable 5 (with fallback)

Anthropic

Intel: 59.9
Speed: 62 t/s
Cost: €8.75 / €43.74
Ctx: 1.0M

State-of-the-art reasoning and complex agentic workflows where budget is no barrier.

#2 · proprietary

GPT-5.6 Sol (max)

OpenAI

Intel: 58.9
Speed: 76 t/s
Cost: €4.37 / €26.25
Ctx: 1.0M

Ultra-high intelligence tasks requiring massive reasoning capabilities and OpenAI ecosystem integration.

#3 · proprietary

Claude Opus 4.8 (max)

Anthropic

Intel: 55.7
Speed: 55 t/s
Cost: €4.37 / €21.87
Ctx: 1.0M

Deep academic research and highly nuanced text analysis at a premium price point.

#4 · proprietary

GPT-5.6 Terra (max)

OpenAI

Intel: 55.0
Speed: 155 t/s
Cost: €2.19 / €13.12
Ctx: 1.0M

High-speed, high-intelligence enterprise automation requiring rapid throughput.

#5 · proprietary

GPT-5.5 (xhigh)

OpenAI

Intel: 54.8
Speed: 81 t/s
Cost: €4.37 / €26.25
Ctx: 922K

Complex reasoning and multi-step problem solving with a large context window.

#6 · proprietary

Grok 4.5 (high)

SpaceXAI

Intel: 53.8
Speed: 114 t/s
Cost: €1.75 / €5.25
Ctx: 500K

Real-time information processing and high-speed reasoning at a competitive price.

#7 · proprietary

Claude Opus 4.7 (max)

Anthropic

Intel: 53.5
Speed: 52 t/s
Cost: €4.37 / €21.87
Ctx: 1.0M

Nuanced literary analysis and complex document synthesis with high reliability.

#8 · proprietary

Claude Sonnet 5 (max)

Anthropic

Intel: 53.4
Speed: 79 t/s
Cost: €1.75 / €8.75
Ctx: 1.0M

Balanced enterprise workflows requiring high intelligence and moderate operational costs.

#9 · proprietary

GPT-5.6 Luna (max)

OpenAI

Intel: 51.2
Speed: 234 t/s
Cost: €0.87 / €5.25
Ctx: 1.0M

High-volume, low-latency customer support and real-time conversational agents.

#10 · open-source

GLM-5.2 (max)

Z AI

Intel: 51.1
Speed: 208 t/s
Cost: €1.22 / €3.85
~€66 / hr*
Ctx: 1.0M

Top-tier open-source deployment for sovereign enterprise applications requiring maximum intelligence.

High-volume customer support
- Blazing speed of 234 tok/s minimises user latency
- Highly economical input cost of €0.87 per Mtok
- Strong intelligence index of 51.2 ensures accurate responses
GPT-5.6 Luna (max)
Sovereign code generation
- Dedicated code-tuned variant with 1000B parameters
- Sovereign open-source deployment with Modified MIT License
- Solid intelligence index of 41.9 for software development
Kimi K2.7 Code
Large document extraction
- Massive context window of 1,048,576 tokens handles huge files
- High intelligence index of 50.6 ensures precise extraction
- Fast processing speed of 130 tok/s accelerates workflows
Muse Spark 1.1 (xhigh)
Long context analysis
- Industry-leading context window of 1,048,576 tokens
- Strong intelligence index of 50.6 for deep synthesis
- Competitive pricing of €1.09 per Mtok input
Muse Spark 1.1 (xhigh)
Sovereign on-premises deployment
- Highest open-source intelligence index of 51.1
- Permissive Mit license allows flexible local deployment
- Efficient 753B parameter architecture balances infra costs
GLM-5.2 (max)
Advanced multimodal vision
- Unmatched intelligence index of 59.9 for complex reasoning
- Generous 1,000,000 token context window for multi-image tasks
- Backed by Anthropic's leading multimodal capabilities
Claude Fable 5 (with fallback)
Complex agentic automation
- Peak intelligence index of 59.9 ensures reliable tool use
- Massive 1,000,000 token context window for long-running agents
- Anthropic's premier tier for complex reasoning tasks
Claude Fable 5 (with fallback)
High-speed multilingual translation
- Blazing speed of 236 tok/s for real-time translation
- Strong intelligence index of 50.2 from Google's translation stack
- Large 1,000,000 token context window for book-length projects
Gemini 3.5 Flash
Structured data extraction
- Massive 1,048,576 token context window for batch processing
- High intelligence index of 50.6 ensures schema adherence
- Fast speed of 130 tok/s delivers rapid structured outputs
Muse Spark 1.1 (xhigh)

How we rank

Snapshot taken 7/14/2026, 2:00:44 AM
Value score = (intelligence − 25)² × min(1, speed ÷ 30) ÷ (input cost + output cost × 3). Output is weighted ×3 because real usage is output-heavy; intelligence is squared so cheap-but-weak models can't dominate.
Intelligence, price and speed data by Artificial Analysis.

Daily refresh at 04:00 Europe/Amsterdam. Numbers move; we don't smooth them.

What is the best AI model right now?

What is the best open-source AI model?

Which AI model has the best price-to-performance?

Which AI model has the largest context window?

How often is the leaderboard updated?

How are the AI models ranked?

Is the SevenLab AI leaderboard free and independent?

A new model just dropped?We track every major release in our daily AI news feed.Read the latest AI news

We pick the model. Then we build the thing.

Building or buying with AI?

Talk directly with our AI specialists

15 min, no strings

No sales pressure

Prototype in 7 days

The AI Leaderboard

Claude Fable 5 (with fallback)

GLM-5.2 (max)

MiMo-V2.5

Top 10, side by side

Claude Fable 5 (with fallback)

GPT-5.6 Sol (max)

Claude Opus 4.8 (max)

GPT-5.6 Terra (max)

GPT-5.5 (xhigh)

Grok 4.5 (high)

Claude Opus 4.7 (max)

Claude Sonnet 5 (max)

GPT-5.6 Luna (max)

GLM-5.2 (max)

Which model for which job

High-volume customer support

Sovereign code generation

Large document extraction

Long context analysis

Sovereign on-premises deployment

Advanced multimodal vision

Complex agentic automation

High-speed multilingual translation

Structured data extraction

How we rank

AI leaderboard FAQ

We pick the model. Then we build the thing.

The AI Leaderboard

Claude Fable 5 (with fallback)

GLM-5.2 (max)

MiMo-V2.5

◈ Top 10, side by side

High-volume customer support

Sovereign code generation

Large document extraction

Long context analysis

Sovereign on-premises deployment

Advanced multimodal vision

Complex agentic automation

High-speed multilingual translation

Structured data extraction

▣ How we rank

AI leaderboard FAQ

We pick the model. Then we build the thing.

Top 10, side by side

How we rank