Model catalog

CosyVoice V3.5 Plus is Alibaba's text-to-speech route for generated voice output. Nfero presents it as part of the audio stack alongside ASR, music, and Qwen Omni models.

Best for: Speech, transcription, and audio workflows

DeepSeek V4 Flash is an Alibaba-hosted partner-model route for faster coding and reasoning tasks. Nfero documents its Token Plan position without claiming a public live demo.

Best for: Coding assistance and developer workflows

DeepSeek V4 Pro is an Alibaba-hosted partner-model route for long-context coding and analytical reasoning. Nfero documents the Token Plan access path without presenting it as a public live demo.

Best for: Coding assistance and developer workflows

Fun ASR is the Alibaba speech-recognition route for turning audio into text. Nfero connects it to transcription, audio understanding, RAG, and agent workflow planning.

Best for: Speech, transcription, and audio workflows

Fun ASR Flash 2026-06-15 is Alibaba's current non-realtime recognition snapshot for multilingual audio files. Nfero records the exact model ID and official access route for implementation planning.

Best for: Speech, transcription, and audio workflows

Fun ASR Realtime is Alibaba's recommended route for continuous multilingual speech recognition and supported dialects. Nfero documents the planning path while public transcription remains disabled.

Best for: Speech, transcription, and audio workflows

Fun Music V1 is the Alibaba text-to-music route in the Nfero audio catalog. It is positioned for creative exploration, music ideation, and audio workflow planning.

Best for: Speech, transcription, and audio workflows

GLM 5 is the baseline GLM route in the Token Plan model set. Nfero keeps it available for teams comparing Zhipu AI options across tool and agent workflows.

Best for: Complex analysis and reasoning tasks

GLM 5.1 is a Zhipu AI Token Plan model for text generation and reasoning. The page helps teams compare GLM routes before choosing a tool or subscription path.

Best for: Complex analysis and reasoning tasks

GLM 5.2 is the Zhipu AI route in Token Plan for reasoning-oriented workflows. Nfero positions it beside Qwen, Kimi, DeepSeek, and MiniMax for buyer comparison.

Best for: Complex analysis and reasoning tasks

HappyHorse 1.0 Video Edit is the earlier editing route for transforming video with prompts and optional reference media. Nfero keeps this page focused on planning and comparison with HappyHorse 1.1.

Best for: Generating and editing video

HappyHorse 1.1 Image-to-Video turns a source image plus optional prompt into generated motion. This page explains the input route and keeps buyers oriented around examples and settings.

Best for: Generating and editing video

HappyHorse 1.1 Reference-to-Video is the route for video work that depends on one or more visual references. Nfero positions it for subject consistency, creative review, and Alibaba access planning.

Best for: Generating and editing video

HappyHorse 1.1 Text-to-Video is Alibaba's high-end prompt-to-video route. Nfero highlights example clips, duration and resolution settings, and planning estimates before approved media access.

Best for: Generating and editing video

Kimi K2.5 is listed for Token Plan reasoning and visual-understanding tasks. It is useful as a partner-model comparison route for teams testing multiple Chinese AI model families.

Best for: Complex analysis and reasoning tasks

Kimi K2.6 is a Moonshot AI Token Plan model for reasoning and multimodal workflows. Nfero uses this page to help teams understand where Kimi fits in the broader Alibaba partner stack.

Best for: Complex analysis and reasoning tasks

Kimi K2.7 Code is the Moonshot AI route in Token Plan for coding-oriented workflows. Nfero pairs it with compatible tools and compares it with Qwen, DeepSeek, GLM, and MiniMax.

Best for: Coding assistance and developer workflows

MiniMax M2.5 is listed in Token Plan for reasoning and text-generation workflows. Nfero presents it as a partner route for teams comparing model behavior across the Alibaba catalog.

Best for: Complex analysis and reasoning tasks

Qwen Audio 3.0 TTS Flash is the low-latency synthesis route for interactive speech and voice-cloning workflows. Alibaba reports time to first audio below 200 ms under supported conditions.

Best for: Speech, transcription, and audio workflows

Qwen Audio 3.0 TTS Plus is the high-quality speech synthesis route for professional voice output and instruction-controlled delivery. Nfero documents its international access and list-price basis without exposing unverified public synthesis.

Best for: Speech, transcription, and audio workflows

Qwen Image is the general Qwen visual generation route for prompt-to-image planning. Use it to understand inputs, aspect ratios, example output, and Alibaba access before generating at scale.

Best for: Creating and editing images

Qwen Image 2.0 is the standard Token Plan Qwen image route for general creative output. The page focuses on prompt input, generated previews, image ratios, and estimated per-image cost.

Best for: Creating and editing images

Qwen Image 2.0 Pro is the higher-capability Qwen image route for polished visuals, product-style images, and creative review. Nfero keeps the prompt box and output gallery visible for real user testing.

Best for: Creating and editing images

Qwen Image 3.0 Pro is Alibaba's limited-preview generation and editing model for complex layouts, small text, detailed photography and multilingual visual content. It accepts up to three reference images and currently requires Model Gallery approval.

Best for: Creating and editing images

Qwen3 Rerank is the Alibaba reranking route for improving retrieval quality after initial search or embedding recall. Nfero connects it to Search and RAG workflows.

Best for: Search and retrieval-augmented generation

Qwen3-VL Embedding converts mixed text, image, and video inputs into configurable vectors for multimodal search and RAG. Nfero documents its limits and Alibaba access route without presenting an unverified live endpoint.

Best for: Search and retrieval-augmented generation

Qwen3-VL Rerank scores mixed text, image, and video candidates after retrieval. Nfero positions it as a multimodal search and RAG planning route with current official documentation.

Best for: Search and retrieval-augmented generation

Qwen3.5 LiveTranslate Flash Realtime combines audio and visual context for multilingual translation, with latency as low as 2.8 seconds under supported conditions. Nfero documents the route without claiming public access.

Best for: Speech, transcription, and audio workflows

Qwen3.5 Omni Plus is Alibaba's multimodal route for combining text, image, audio, and video understanding or generation. Nfero places it in the broader Omni and agent-planning surface.

Best for: Speech, transcription, and audio workflows

Qwen3.5 Omni Plus Realtime is the low-latency Omni route for interactive multimodal experiences. Nfero uses it for planning realtime, audio, and agent-facing Alibaba workflows.

Best for: Speech, transcription, and audio workflows

Qwen3.6 Flash is the faster Qwen route for lightweight prompts, tool trials, and rapid comparison against Max and Plus models. It is useful when latency and demo volume matter.

Best for: Coding assistance and developer workflows

Qwen3.6 Plus is listed for Token Plan workflows that combine text generation, reasoning, and visual understanding. Use this page to plan compatible tool access and compare it with newer Qwen routes.

Best for: Complex analysis and reasoning tasks

Qwen3.7 Max is the premium Qwen route on Nfero for reasoning, coding, and executive-quality answer testing. Use the live demo to validate output quality and cost before continuing through Alibaba Cloud or Token Plan.

Best for: Coding assistance and developer workflows

Qwen3.7 Plus is positioned as a balanced Qwen option for teams that need strong reasoning without always using the highest-cost route. Nfero shows live output, latency, and estimated usage cost.

Best for: Coding assistance and developer workflows

Qwen3.8 Max is Alibaba's production flagship for long-context reasoning, coding, visual understanding and text generation. Nfero shows its official limits, current access routes and Singapore pay-as-you-go rates.

Best for: Coding assistance and developer workflows

Text Embedding V4 is Alibaba's text embedding route for semantic indexing and retrieval. Nfero uses it as the foundation page for Search and RAG planning.

Best for: Search and retrieval-augmented generation

Tongyi Embedding Vision Plus is Alibaba's multimodal embedding route for image and text retrieval. Nfero positions it for visual search, catalog retrieval, and grounded multimodal applications.

Best for: Search and retrieval-augmented generation

Wan 2.7 Image is listed as an Alibaba Token Plan image-generation model. Nfero presents it as a visual route for teams comparing Wan and Qwen image models.

Best for: Creating and editing images

Wan 2.7 Image Pro is the pro Wan image route in the Token Plan catalog. The page is designed for buyers comparing image quality, settings, and Alibaba continuation paths.

Best for: Creating and editing images

Wan 2.7 Image-to-Video turns source imagery into 2-15 second 720P or 1080P clips. Nfero documents its main workflows and current Alibaba Model Studio route without enabling unverified public generation.

Best for: Generating and editing video

Wan 2.7 Reference-to-Video uses visual references to preserve subjects and appearance across generated clips. Nfero keeps its current limits and Alibaba access guidance visible for production planning.

Best for: Generating and editing video

Wan 2.7 Text-to-Video is Alibaba's current prompt-to-video route for multi-shot 720P and 1080P clips with synchronized audio. Nfero explains the supported controls and official access path without claiming a public demo or Token Plan inclusion.

Best for: Generating and editing video

Wan 2.7 Video Edit is Alibaba's instruction-guided route for controlled video transformations and style transfer. Nfero documents the workflow while keeping unverified pricing and public execution disabled.

Best for: Generating and editing video

Wan Text-to-Video Turbo is the fast Wan route for turning prompts into short video drafts. Nfero focuses on safe planning, settings, and examples before production media access.

Best for: Generating and editing video