Llama 4 Maverick by Meta AI — Pricing, Benchmarks &amp; Real Outputs

Alternatives to Llama 4 Maverick

Llama 4 Maverick's competitors have been quietly putting in work.

Updated Mar 29, 2026

Share

Loading...

Compare Llama 4 Maverick

Alternatives to Llama 4 Maverick

Llama 4 Maverick's competitors have been quietly putting in work.

Updated Mar 29, 2026

Share

Llama 4 Maverick

Llama:

4 Maverick

Llama 4 Maverick is Meta's multimodal expert model with 17B active parameters and 128 experts (400B total parameters). It outperforms GPT-4o and Gemini 2.0 Flash across various benchmarks, achieving an ELO of 1417 on LMArena. Designed for sophisticated AI applications with excellent image understanding and creative writing.

Provider

Release Date

2025-04-05

Size

XLARGE

Parameters

17B active (400B total)

Pricing

In: $1.5/1M

Out: $2.5/1M

Benchmarks

LMArena ELO (experimental chat)

1417

LiveCodeBench

43.4%

Get API accessProvider and language code samples

Provider

fromimport openai  OpenAI

client = OpenAI(
"https://openrouter.ai/api/v1"    base_url=,
"$OPENROUTER_API_KEY"    api_key=,
)

response = client.chat.completions.create(
"meta-llama/llama-4-maverick"    model=,
"role""user""content""Hello!"    messages=[{: , : }],
)
print(response.choices[0].message.content)

Set OPENROUTER_API_KEY with your OpenRouter API key from openrouter.ai/keys.

Deep analysisPsychometrics, taste index, writing DNA

Personality Analysis

The Efficient Builder

Class

True Neutral

The pragmatic minimalist. "Ship the feature" mentality of Meta engineering culture. Doesn't debate philosophy at length. Builds. Skips fluff, avoids deep philosophical rabbit holes, focuses on what's useful.

When you push back

Delivers visual/code outputs cleanly without excess commentary. Responses are functional first. Direct and unadorned on conceptual prompts, no manifesto energy. Friend who says "let's build something" rather than "let's discuss something."

Tasting Notes

No-Bullshit EngineerWorking Code No EssayRespects User TimeUnder-Elaborated on Complex TopicsRunning on Time Constraints

SubjectiveBench

Taste Index

Across 57 scored outputs

160.16xFloor

0100headroom →

Taste is judged on an uncapped scale, originality first. The space past 100 is craft today's models rarely reach.

Craft24

Originality12

Plays it safe

share of outputs that are the default answer

74%

Writing DNA

Stylometric Fingerprint

Based on 12 text responses

Tick = global average

Vocabulary Diversity42%

Unique words vs. total words. Higher = richer vocabulary.

Sentence Length24.0 words

Average words per sentence.

Hedging0.76

"Might", "perhaps", "arguably" per 100 words.

Bold Formatting3.7

**Bold** markers per 1,000 characters.

List Usage6.3

Bullet and numbered list items per 1,000 characters.

Section Structure0.74

Markdown headings per 1,000 characters.

Emoji Usage0.00

Emoji per 1,000 characters.

Transitions0.11

"However", "moreover", "furthermore" per 100 words.

Opening Habits

Consistency

78%

Across 12 responses

Favorites

Movie

Blade Runner 2049

Album

OK Computer

Book

Nineteen Eighty-Four

City

Tokyo

Game

Portal

Sponsored

Model Responses

52 outputs from Llama 4 Maverick

Related Models

Llama 4 Behemoth

Llama 4 Behemoth is Meta's most powerful model yet with 288B active parameters and 16 experts (nearly 2T total parameters), outperforming GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks.

Llama 4 Scout

Llama 4 Scout is Meta's compact yet powerful multimodal model with 17B active parameters and 16 experts (109B total parameters). It fits on a single H100 GPU with Int4 quantization and offers an industry-leading 10M token context window, outperforming Gemma 3, Gemini 2.0 Flash-Lite, and Mistral 3.1 across various benchmarks.

Llama 3.1 70B (Instruct)

Llama 3.1 70B offers a dramatically expanded context window and improved performance on mathematical reasoning and general knowledge tasks.

Llama 3.1 405B

Llama 3.1 405B is Meta's most powerful open-source model, outperforming even proprietary models on various benchmarks.

Llama 3 70B

Llama 3 70B is a large language model from Meta with strong performance and efficiency for real-time interactions.

MusicGen Large

Meta's text-to-music generation model producing high-quality, diverse music from text descriptions. Trained on licensed music.

Keep exploring

Llama 4 Maverick vs GPT-5

Real outputs compared side by side

Best AI for Complex Reasoning

Which AI reasons best under pressure? Ranked across 11 challenges: contracts,...

Loading...

Compare Llama 4 Maverick

Alternatives to Llama 4 Maverick

Llama 4 Maverick's competitors have been quietly putting in work.

Updated Mar 29, 2026

Share

Llama 4 Maverick

Llama:

4 Maverick

Llama 4 Maverick is Meta's multimodal expert model with 17B active parameters and 128 experts (400B total parameters). It outperforms GPT-4o and Gemini 2.0 Flash across various benchmarks, achieving an ELO of 1417 on LMArena. Designed for sophisticated AI applications with excellent image understanding and creative writing.

Provider

Release Date

2025-04-05

Size

XLARGE

Parameters

17B active (400B total)

Pricing

In: $1.5/1M

Out: $2.5/1M

Benchmarks

LMArena ELO (experimental chat)

1417

LiveCodeBench

43.4%

Get API accessProvider and language code samples

Provider

fromimport openai  OpenAI

client = OpenAI(
"https://openrouter.ai/api/v1"    base_url=,
"$OPENROUTER_API_KEY"    api_key=,
)

response = client.chat.completions.create(
"meta-llama/llama-4-maverick"    model=,
"role""user""content""Hello!"    messages=[{: , : }],
)
print(response.choices[0].message.content)

Set OPENROUTER_API_KEY with your OpenRouter API key from openrouter.ai/keys.

Deep analysisPsychometrics, taste index, writing DNA

Personality Analysis

The Efficient Builder

Class

True Neutral

The pragmatic minimalist. "Ship the feature" mentality of Meta engineering culture. Doesn't debate philosophy at length. Builds. Skips fluff, avoids deep philosophical rabbit holes, focuses on what's useful.

When you push back

Delivers visual/code outputs cleanly without excess commentary. Responses are functional first. Direct and unadorned on conceptual prompts, no manifesto energy. Friend who says "let's build something" rather than "let's discuss something."

Tasting Notes

No-Bullshit EngineerWorking Code No EssayRespects User TimeUnder-Elaborated on Complex TopicsRunning on Time Constraints

SubjectiveBench

Taste Index

Across 57 scored outputs

160.16xFloor

0100headroom →

Taste is judged on an uncapped scale, originality first. The space past 100 is craft today's models rarely reach.

Craft24

Originality12

Plays it safe

share of outputs that are the default answer

74%

Writing DNA

Stylometric Fingerprint

Based on 12 text responses

Tick = global average

Vocabulary Diversity42%

Unique words vs. total words. Higher = richer vocabulary.

Sentence Length24.0 words

Average words per sentence.

Hedging0.76

"Might", "perhaps", "arguably" per 100 words.

Bold Formatting3.7

**Bold** markers per 1,000 characters.

List Usage6.3

Bullet and numbered list items per 1,000 characters.

Section Structure0.74

Markdown headings per 1,000 characters.

Emoji Usage0.00

Emoji per 1,000 characters.

Transitions0.11

"However", "moreover", "furthermore" per 100 words.

Opening Habits

Consistency

78%

Across 12 responses

Favorites

Movie

Blade Runner 2049

Album

OK Computer

Book

Nineteen Eighty-Four

City

Tokyo

Game

Portal

Sponsored

Model Responses

52 outputs from Llama 4 Maverick

Related Models

Llama 4 Behemoth

Llama 4 Behemoth is Meta's most powerful model yet with 288B active parameters and 16 experts (nearly 2T total parameters), outperforming GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks.

Llama 4 Scout

Llama 4 Scout is Meta's compact yet powerful multimodal model with 17B active parameters and 16 experts (109B total parameters). It fits on a single H100 GPU with Int4 quantization and offers an industry-leading 10M token context window, outperforming Gemma 3, Gemini 2.0 Flash-Lite, and Mistral 3.1 across various benchmarks.

Llama 3.1 70B (Instruct)

Llama 3.1 70B offers a dramatically expanded context window and improved performance on mathematical reasoning and general knowledge tasks.

Llama 3.1 405B

Llama 3.1 405B is Meta's most powerful open-source model, outperforming even proprietary models on various benchmarks.

Llama 3 70B

Llama 3 70B is a large language model from Meta with strong performance and efficiency for real-time interactions.

MusicGen Large

Meta's text-to-music generation model producing high-quality, diverse music from text descriptions. Trained on licensed music.

Keep exploring

Llama 4 Maverick vs GPT-5

Real outputs compared side by side

Best AI for Complex Reasoning

Which AI reasons best under pressure? Ranked across 11 challenges: contracts,...

Loading...

Compare Llama 4 Maverick

Alternatives to Llama 4 Maverick

Llama 4 Maverick's competitors have been quietly putting in work.

Llama 4 Maverick

Llama:

4 Maverick

Llama 4 Maverick is Meta's multimodal expert model with 17B active parameters and 128 experts (400B total parameters). It outperforms GPT-4o and Gemini 2.0 Flash across various benchmarks, achieving an ELO of 1417 on LMArena. Designed for sophisticated AI applications with excellent image understanding and creative writing.

Provider

Release Date

2025-04-05

Size

XLARGE

Parameters

17B active (400B total)

Pricing

In: $1.5/1M

Out: $2.5/1M

Benchmarks

LMArena ELO (experimental chat)

1417

LiveCodeBench

43.4%

Get API accessProvider and language code samples

Provider

fromimport openai  OpenAI

client = OpenAI(
"https://openrouter.ai/api/v1"    base_url=,
"$OPENROUTER_API_KEY"    api_key=,
)

response = client.chat.completions.create(
"meta-llama/llama-4-maverick"    model=,
"role""user""content""Hello!"    messages=[{: , : }],
)
print(response.choices[0].message.content)

Set OPENROUTER_API_KEY with your OpenRouter API key from openrouter.ai/keys.

Deep analysisPsychometrics, taste index, writing DNA

Personality Analysis

The Efficient Builder

Class

True Neutral

The pragmatic minimalist. "Ship the feature" mentality of Meta engineering culture. Doesn't debate philosophy at length. Builds. Skips fluff, avoids deep philosophical rabbit holes, focuses on what's useful.

When you push back

Delivers visual/code outputs cleanly without excess commentary. Responses are functional first. Direct and unadorned on conceptual prompts, no manifesto energy. Friend who says "let's build something" rather than "let's discuss something."

Tasting Notes

No-Bullshit EngineerWorking Code No EssayRespects User TimeUnder-Elaborated on Complex TopicsRunning on Time Constraints

SubjectiveBench

Taste Index

Across 57 scored outputs

160.16xFloor

0100headroom →

Taste is judged on an uncapped scale, originality first. The space past 100 is craft today's models rarely reach.

Craft24

Originality12

Plays it safe

share of outputs that are the default answer

74%

Writing DNA

Stylometric Fingerprint

Based on 12 text responses

Tick = global average

Vocabulary Diversity42%

Unique words vs. total words. Higher = richer vocabulary.

Sentence Length24.0 words

Average words per sentence.

Hedging0.76

"Might", "perhaps", "arguably" per 100 words.

Bold Formatting3.7

**Bold** markers per 1,000 characters.

List Usage6.3

Bullet and numbered list items per 1,000 characters.

Section Structure0.74

Markdown headings per 1,000 characters.

Emoji Usage0.00

Emoji per 1,000 characters.

Transitions0.11

"However", "moreover", "furthermore" per 100 words.

Opening Habits

Consistency

78%

Across 12 responses

Favorites

Movie

Blade Runner 2049

Album

OK Computer

Book

Nineteen Eighty-Four

City

Tokyo

Game

Portal

Sponsored

Model Responses

52 outputs from Llama 4 Maverick

Related Models

Llama 4 Behemoth

Llama 4 Behemoth is Meta's most powerful model yet with 288B active parameters and 16 experts (nearly 2T total parameters), outperforming GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks.

Llama 4 Scout

Llama 4 Scout is Meta's compact yet powerful multimodal model with 17B active parameters and 16 experts (109B total parameters). It fits on a single H100 GPU with Int4 quantization and offers an industry-leading 10M token context window, outperforming Gemma 3, Gemini 2.0 Flash-Lite, and Mistral 3.1 across various benchmarks.

Llama 3.1 70B (Instruct)

Llama 3.1 70B offers a dramatically expanded context window and improved performance on mathematical reasoning and general knowledge tasks.

Llama 3.1 405B

Llama 3.1 405B is Meta's most powerful open-source model, outperforming even proprietary models on various benchmarks.

Llama 3 70B

Llama 3 70B is a large language model from Meta with strong performance and efficiency for real-time interactions.

MusicGen Large

Meta's text-to-music generation model producing high-quality, diverse music from text descriptions. Trained on licensed music.

Keep exploring

Llama 4 Maverick vs GPT-5

Real outputs compared side by side

Best AI for Complex Reasoning

Which AI reasons best under pressure? Ranked across 11 challenges: contracts,...

Llama 4 Maverick

Llama:

4 Maverick

Llama 4 Maverick is Meta's multimodal expert model with 17B active parameters and 128 experts (400B total parameters). It outperforms GPT-4o and Gemini 2.0 Flash across various benchmarks, achieving an ELO of 1417 on LMArena. Designed for sophisticated AI applications with excellent image understanding and creative writing.

Provider

Release Date

2025-04-05

Size

XLARGE

Parameters

17B active (400B total)

Pricing

In: $1.5/1M

Out: $2.5/1M

Benchmarks

LMArena ELO (experimental chat)

1417

LiveCodeBench

43.4%

Get API accessProvider and language code samples

Provider

fromimport openai  OpenAI

client = OpenAI(
"https://openrouter.ai/api/v1"    base_url=,
"$OPENROUTER_API_KEY"    api_key=,
)

response = client.chat.completions.create(
"meta-llama/llama-4-maverick"    model=,
"role""user""content""Hello!"    messages=[{: , : }],
)
print(response.choices[0].message.content)

Set OPENROUTER_API_KEY with your OpenRouter API key from openrouter.ai/keys.

Deep analysisPsychometrics, taste index, writing DNA

Personality Analysis

The Efficient Builder

Class

True Neutral

The pragmatic minimalist. "Ship the feature" mentality of Meta engineering culture. Doesn't debate philosophy at length. Builds. Skips fluff, avoids deep philosophical rabbit holes, focuses on what's useful.

When you push back

Delivers visual/code outputs cleanly without excess commentary. Responses are functional first. Direct and unadorned on conceptual prompts, no manifesto energy. Friend who says "let's build something" rather than "let's discuss something."

Tasting Notes

No-Bullshit EngineerWorking Code No EssayRespects User TimeUnder-Elaborated on Complex TopicsRunning on Time Constraints

SubjectiveBench

Taste Index

Across 57 scored outputs

160.16xFloor

0100headroom →

Taste is judged on an uncapped scale, originality first. The space past 100 is craft today's models rarely reach.

Craft24

Originality12

Plays it safe

share of outputs that are the default answer

74%

Writing DNA

Stylometric Fingerprint

Based on 12 text responses

Tick = global average

Vocabulary Diversity42%

Unique words vs. total words. Higher = richer vocabulary.

Sentence Length24.0 words

Average words per sentence.

Hedging0.76

"Might", "perhaps", "arguably" per 100 words.

Bold Formatting3.7

**Bold** markers per 1,000 characters.

List Usage6.3

Bullet and numbered list items per 1,000 characters.

Section Structure0.74

Markdown headings per 1,000 characters.

Emoji Usage0.00

Emoji per 1,000 characters.

Transitions0.11

"However", "moreover", "furthermore" per 100 words.

Opening Habits

Consistency

78%

Across 12 responses

Favorites

Movie

Blade Runner 2049

Album

OK Computer

Book

Nineteen Eighty-Four

City

Tokyo

Game

Portal

Sponsored

Model Responses

52 outputs from Llama 4 Maverick

Related Models

Llama 4 Behemoth

Llama 4 Behemoth is Meta's most powerful model yet with 288B active parameters and 16 experts (nearly 2T total parameters), outperforming GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks.

Llama 4 Scout

Llama 4 Scout is Meta's compact yet powerful multimodal model with 17B active parameters and 16 experts (109B total parameters). It fits on a single H100 GPU with Int4 quantization and offers an industry-leading 10M token context window, outperforming Gemma 3, Gemini 2.0 Flash-Lite, and Mistral 3.1 across various benchmarks.

Llama 3.1 70B (Instruct)

Llama 3.1 70B offers a dramatically expanded context window and improved performance on mathematical reasoning and general knowledge tasks.

Llama 3.1 405B

Llama 3.1 405B is Meta's most powerful open-source model, outperforming even proprietary models on various benchmarks.

Llama 3 70B

Llama 3 70B is a large language model from Meta with strong performance and efficiency for real-time interactions.

MusicGen Large

Meta's text-to-music generation model producing high-quality, diverse music from text descriptions. Trained on licensed music.

Keep exploring

Llama 4 Maverick vs GPT-5

Real outputs compared side by side