NEWSROOM
News & Announcements
Partnerships, product launches, research and company news from the MythWorx team.
Product
3 min read
Introducing NeuroWorx℠: AI reasoning at 2% of the power
NeuroWorx℠ is a verifier-first reasoning engine that runs on CPUs, explains every conclusion and uses about 2% of the power most LLMs need.
Company
4 min read
Jason Williamson named CEO as MythWorx outperforms hundreds of LLMs on MMLU-Pro
The former Oracle executive and U.S. Marine Corps veteran takes the helm as MythWorx’s hybrid architecture posts standout MMLU-Pro results at a fraction of the power.
Research
3 min read
Third Eye Advisory validates MythWorx’s human-like approach to AI
An independent analyst report finds MythWorx’s smaller, reasoning-based architecture a credible alternative to the brute-force scale of traditional LLMs.
Benchmark
2 min read
MythWorx scores 71.24% on MMLU-Pro with zero pretraining
In a single zero-shot attempt across more than 12,000 tasks, MythWorx’s 14B-parameter model beat models hundreds of times larger, using a fraction of the compute.
Funding
3 min read
MythWorx closes $5M seed round at a $100M valuation
Led by Eagle Venture Fund IV and Eagle Freedom Funds I and II, the round will fund hiring, new products and a faster go-to-market.
Research
3 min read
Capability white paper: reasoning with a tenth of the infrastructure
The ideas behind MythWorx’s brain-inspired AI: how it learns, where it applies, and why it needs far less compute than LLMs.
Product
3 min read
Introducing NeuroWorx℠: AI reasoning at 2% of the power
NeuroWorx℠ is a verifier-first reasoning engine that runs on CPUs, explains every conclusion and uses about 2% of the power most LLMs need.
Funding
3 min read
MythWorx closes $5M seed round at a $100M valuation
Led by Eagle Venture Fund IV and Eagle Freedom Funds I and II, the round will fund hiring, new products and a faster go-to-market.
Benchmark
2 min read
MythWorx scores 71.24% on MMLU-Pro with zero pretraining
In a single zero-shot attempt across more than 12,000 tasks, MythWorx’s 14B-parameter model beat models hundreds of times larger, using a fraction of the compute.
Research
3 min read
Third Eye Advisory validates MythWorx’s human-like approach to AI
An independent analyst report finds MythWorx’s smaller, reasoning-based architecture a credible alternative to the brute-force scale of traditional LLMs.
Company
4 min read
Jason Williamson named CEO as MythWorx outperforms hundreds of LLMs on MMLU-Pro
The former Oracle executive and U.S. Marine Corps veteran takes the helm as MythWorx’s hybrid architecture posts standout MMLU-Pro results at a fraction of the power.
Research
3 min read
Capability white paper: reasoning with a tenth of the infrastructure
The ideas behind MythWorx’s brain-inspired AI: how it learns, where it applies, and why it needs far less compute than LLMs.
Product
3 min read
Introducing NeuroWorx℠: AI reasoning at 2% of the power
NeuroWorx℠ is a verifier-first reasoning engine that runs on CPUs, explains every conclusion and uses about 2% of the power most LLMs need.
Research
3 min read
Third Eye Advisory validates MythWorx’s human-like approach to AI
An independent analyst report finds MythWorx’s smaller, reasoning-based architecture a credible alternative to the brute-force scale of traditional LLMs.
Funding
3 min read
MythWorx closes $5M seed round at a $100M valuation
Led by Eagle Venture Fund IV and Eagle Freedom Funds I and II, the round will fund hiring, new products and a faster go-to-market.
Company
4 min read
Jason Williamson named CEO as MythWorx outperforms hundreds of LLMs on MMLU-Pro
The former Oracle executive and U.S. Marine Corps veteran takes the helm as MythWorx’s hybrid architecture posts standout MMLU-Pro results at a fraction of the power.
Benchmark
2 min read
MythWorx scores 71.24% on MMLU-Pro with zero pretraining
In a single zero-shot attempt across more than 12,000 tasks, MythWorx’s 14B-parameter model beat models hundreds of times larger, using a fraction of the compute.
Research
3 min read
Capability white paper: reasoning with a tenth of the infrastructure
The ideas behind MythWorx’s brain-inspired AI: how it learns, where it applies, and why it needs far less compute than LLMs.
Product
3 min read
Introducing NeuroWorx℠: AI reasoning at 2% of the power
NeuroWorx℠ is a verifier-first reasoning engine that runs on CPUs, explains every conclusion and uses about 2% of the power most LLMs need.
Benchmark
2 min read
MythWorx scores 71.24% on MMLU-Pro with zero pretraining
In a single zero-shot attempt across more than 12,000 tasks, MythWorx’s 14B-parameter model beat models hundreds of times larger, using a fraction of the compute.
Research
3 min read
Third Eye Advisory validates MythWorx’s human-like approach to AI
An independent analyst report finds MythWorx’s smaller, reasoning-based architecture a credible alternative to the brute-force scale of traditional LLMs.
Research
3 min read
Capability white paper: reasoning with a tenth of the infrastructure
The ideas behind MythWorx’s brain-inspired AI: how it learns, where it applies, and why it needs far less compute than LLMs.
You already built the foundation.
Let’s help it go further.
Whether it’s a factory floor, a ship, a hospital, or an air-gapped facility — if your AI needs to work where the cloud can’t follow, we should talk.
You already built the foundation. Let’s help it go further.
Whether it’s a factory floor, a ship, a hospital, or an air-gapped facility — if your AI needs to work where the cloud can’t follow, we should talk.
