SuperAPI AI Directory

All AI tools

35,738 results · Page 305 of 993

mixedvoices

AI developer platforms

MixedVoices is an analysis and evaluation tool for voice agents, positioned similarly to "Mixpanel for conversational AI." It tracks, visualizes, and optimizes agent performance through conversation analysis, call quality metrics, and call flow diagrams, and supports running simulation tests before launching in production. Core capabilities include: quick integration via Python API, interactive dashboards, call flow analysis (showing conversation paths, patterns, and success rates), version control (comparing agent behavior across iterations), machine learning performance metrics (hallucination, call scheduling, conciseness, empathy scores), call quality analysis (interruptions, latency, signal-to-noise ratio, words per minute), and agent evaluation through simulation and stress testing. Before use, models must be configured via mixedvoices config. By default, OpenAI models are used for analysis and transcription. Analysis supports gpt-3.5 and later GPT models, and transcription supports OpenAI whisper and Deepgram nova-2. The OPENAI_API_KEY and DEEPGRAM_API_KEY environment variables must be set accordingly. The tool supports analyzing recordings via Python API, evaluating custom agents and Bland AI Agent, and launching an interactive dashboard from the command line. The project uses the Apache 2.0 license.

Read more

ML.ai

AI coding

ML.ai is an AI coding agent that can read code repositories, route each step of a task to the appropriate model, and show a diff before any changes are written to disk, requiring the user to click Allow. It can handle bug fixes, test suites, or week-long migration tasks, automatically selecting models, running tests, and opening PRs. Users can operate it from the terminal, IDE, GitHub, or the web. It supports macOS, Linux, and Windows, and is currently launching first in India.

Read more

MLA header generator

AI search

MLA Header Generator is a free academic formatting tool that helps users quickly generate an MLA-formatted paper header. It provides formatting examples, explanations of common mistakes, and a live preview. Users can fill in five pieces of information—name, teacher, course, date, and title—without registering, and generate a downloadable DOCX file or copy the text to Google Docs. The tool also includes a formatting checklist to help users avoid common typesetting errors.

Read more

MMR1

AI developer platforms

MMR1 is an open-source research project for multimodal reasoning models (CVPR 2026), developed by LengSicong et al. The project addresses two major bottlenecks currently facing multimodal reasoning training—the lack of open, large-scale, high-quality long chain-of-thought (CoT) data, and the susceptibility of standard GRPO reinforcement learning to vanishing gradients under low reward variance, which leads to unstable fine-tuning—by proposing the Variance-Aware Sampling (VAS) method. Guided by the Variance Promotion Score (VPS), VAS combines outcome variance and trajectory diversity to select training prompts, thereby increasing reward variance, stabilizing policy optimization, and accelerating convergence. The project also open-sources the MMR1-SFT dataset of approximately 1.6 million long CoT cold-start trajectories, the MMR1-RL dataset of approximately 15,000 RL QA pairs, and multi-scale model checkpoints at 3B, 7B, and 32B, along with an end-to-end reproducible training pipeline. Evaluations show that MMR1-7B-RL achieves 7B-scale SOTA on multiple mathematical multimodal reasoning benchmarks. The project is licensed under Apache 2.0 and is a research preview for non-commercial use only.

Read more