What other models does Alibaba offer?

Qwen3.5 Plus is one API tier in Alibaba's Qwen lineup, which spans lightweight flash models, large-scale open-weight reasoning models, coding specialists, and multimodal offerings. Here are some other models Alibaba supports:

• Qwen3.7 Max: Qwen3.7 Max is Alibaba's current proprietary flagship reasoning model, scoring 57 on the Artificial Analysis Intelligence Index (as of 06/04/2026) and ranking in the top 10 of over 150 evaluated models, with a 1M-token context window and 182.7 tokens per second output speed at $2.50 per 1M input and $7.50 per 1M output

• Qwen3.6 Plus: Qwen3.6 Plus is a proprietary mid-tier reasoning model from Alibaba with a 1M-token context window and multimodal input support, scoring 50 on the Intelligence Index (as of 06/04/2026) at $0.50 per 1M input and $3.00 per 1M output, released April 2026

• Qwen3.5 Flash: Qwen3.5 Flash is the faster, lower-cost sibling in the Qwen3.5 API tier, built on the 122B A10B open-weight model, scoring 42 on the Intelligence Index (as of 06/04/2026) and generating 140.6 tokens per second at $0.40 per 1M input and $3.20 per 1M output

• Qwen3 Coder Next: Qwen3 Coder Next is Alibaba's open-weight coding-specialized model at 79.7B total parameters with a 256k-token context window, optimized for code generation and agentic programming tasks at $0.35 per 1M input and $1.20 per 1M output under Apache 2.0

• Qwen3.5 Omni Plus: Qwen3.5 Omni Plus is a fully multimodal model from Alibaba accepting text, image, speech, and video as input and producing text and speech as output, with a 256k-token context window, designed for voice-enabled and video understanding workflows at $0.40 per 1M input and $4.80 per 1M output

How does Qwen3.5 Plus differ from Alibaba's other models?

Qwen3.5 Plus is the larger of the two Qwen3.5 API tiers, running on the underlying Qwen3.5 397B A17B open-weight MoE model, and is positioned as the high-capability option in the Qwen3.5 generation before the proprietary Qwen3.6 and Qwen3.7 tiers.

• Intelligence Index: Qwen3.5 Plus scores 45 on the Artificial Analysis Intelligence Index (as of 06/04/2026), ranking 13th of 89 comparable models. This places it above the Qwen3.5 Flash tier (score: 42) and the earlier Qwen3 Next generation, while trailing the proprietary Qwen3.6 Plus (score: 50) and Qwen3.7 Max (score: 57)

• Pricing: At $0.60 per 1M input and $3.60 per 1M output, Qwen3.5 Plus costs half as much on input as Qwen3.6 Plus ($0.50 input, $3.00 output), though output costs are comparable. Compared to the proprietary Qwen3.7 Max at $7.50 per 1M output, Qwen3.5 Plus offers meaningfully lower output cost for teams that don't need flagship-tier reasoning

• Speed: Qwen3.5 Plus generates 52.2 tokens per second (as of 06/04/2026), which is slower than the Qwen3.5 Flash tier at 140.6 tokens per second. For latency-sensitive applications at scale, the Flash tier is the better fit, while Plus is appropriate when throughput is less critical than benchmark quality

• Context window: Both Qwen3.5 Plus and Qwen3.5 Flash offer a 262k-token context window, consistent with the open-weight MoE models they're built on. This is shorter than the 1M-token windows available on Qwen3.6 Plus and Qwen3.7 Max, which matters for very long document or multi-turn workloads

• Modalities: Qwen3.5 Plus accepts text and image as input and produces text only. It doesn't support audio or video, unlike the Qwen3.5 Omni series, which limits it to text and vision-based task pipelines

Qwen3.5 Plus is the right choice when teams need a reasoning-capable model with strong benchmark performance and open-weight transparency at a cost tier well below the Qwen3.6 and Qwen3.7 proprietary flagships. It's worth noting that "Plus" is the API product name and the underlying weights are Qwen3.5 397B A17B, which can be self-hosted under Apache 2.0.

What models should I consider using alongside Qwen3.5 Plus?

No single model is optimal for every task. Here are models worth pairing with Qwen3.5 Plus depending on what your product needs:

• Qwen3.5 Flash (Alibaba): For high-volume, lower-complexity requests where 140.6 tokens per second throughput matters and the Intelligence Index score of 42 is sufficient, routing to Qwen3.5 Flash reduces per-token cost while staying within the same model family and context window specification

• Qwen3.7 Max (Alibaba): When a task requires the highest reasoning quality Alibaba offers, such as complex multi-step analysis or long agentic chains, routing to Qwen3.7 Max with its Intelligence Index score of 57 (as of 06/04/2026) is appropriate when the premium output cost is justified

• Claude Sonnet 4.5 (Anthropic): For structured output generation, strict instruction adherence, or production workflows where cross-provider redundancy is important, Claude Sonnet 4.5 provides a well-documented alternative that complements Qwen3.5 Plus as a failover or for specific instruction-heavy task types

• Gemini 2.0 Flash (Google): When requests include audio or video content that Qwen3.5 Plus can't process, routing to Gemini 2.0 Flash adds multimodal coverage at low cost and high throughput without restructuring the core pipeline

• GPT-4.1 Mini (OpenAI): For simple classification, extraction, or short-form summarization tasks where Qwen3.5 Plus has more capacity than the task demands, GPT-4.1 Mini offers broad regional availability and a lower blended cost per token across high-volume simple workloads

What are the challenges of using Qwen3.5 Plus in my product?

Like any production LLM, Qwen3.5 Plus comes with tradeoffs worth planning for:

• Throughput ceiling: At 52.2 tokens per second (as of 06/04/2026), Qwen3.5 Plus is one of the slower models in the Qwen3.5 lineup. Applications with latency budgets under one second for typical completions, or high-concurrency requirements, should evaluate whether the Flash tier or a different provider better fits the speed profile

• Context window vs. proprietary tiers: The 262k-token context window is solid for most document tasks, but if your application needs the 1M-token capacity available on Qwen3.6 Plus or Qwen3.7 Max, Qwen3.5 Plus can't substitute. This creates a hard ceiling for very long retrieval or multi-document synthesis pipelines

• Text and image input only: Qwen3.5 Plus doesn't support audio or video inputs, which means any pipeline receiving multimodal content beyond images needs separate routing logic for those request types, adding branching complexity

• Provider dependency: Relying on Alibaba's DashScope API as a single provider creates fragility when the provider has an outage or deprecates a model version. Alibaba releases successive Qwen generations quickly, so monitoring for model ID deprecations is part of ongoing maintenance

• Cost at scale: As request volume grows, token costs compound quickly without active cost management. At $3.60 per 1M output tokens, workloads with long completions or high concurrency will accumulate costs substantially faster than lower-output-cost alternatives in the Qwen family

Why should I use Merge Gateway to route LLM requests with Qwen3.5 Plus and every other model?

Using Qwen3.5 Plus through Merge Gateway gives you access to the model itself and the infrastructure layer around it:

• Intelligent routing and automatic failover: Merge routes around Alibaba outages automatically. Routing policies based on cost, latency, or quality can reduce spend by 40 to 60% without touching your application code, and you can define Qwen3.5 Flash as a cost-based fallback within the same policy

• One API, every provider: Access Qwen3.5 Plus and every other major LLM through a single endpoint and API key. Change providers by swapping the model string, with no application code changes required

• Cost governance: Set hard or soft project budgets so Qwen3.5 Plus spend stays within plan. Every request is attributed to a model, project, and tag in a unified billing dashboard across all providers

• Build Your Own Router: Define what "best" means for your traffic by selecting from curated ML benchmarks or adding your own eval scores. The router scores each available model against your weights and picks the winner per request, with a plain-language explanation of every decision

• Security and compliance controls: Apply DLP rules and prompt injection protection before every request reaches Alibaba. Enforce per-project model and region policies without adding that logic to your application

How can I start routing requests to Qwen3.5 Plus via Merge Gateway?

Getting Qwen3.5 Plus running through Merge Gateway takes a few minutes:

1. Create an account and get your API key from the dashboard.

2. Install the Merge Gateway SDK: run pip install merge-gateway-sdk (Python) or npm install merge-gateway-sdk (Node). Alternatively, if you're already using the OpenAI SDK, set base_url = "https://api-gateway.merge.dev/v1/openai" and your existing code works as-is.

3. Make your first request using the provider/model format. For Qwen3.5 Plus, the model string is alibaba/qwen3.5-plus. Swap the model string to route to any other provider without changing anything else.

4. Configure a routing policy in the dashboard to set failover behavior, cost limits, and optimization strategy. Your first policy can be as simple as naming Qwen3.5 Plus as primary with one fallback.

Full setup instructions and SDK references are in the Merge Gateway docs.