LLM System Design and Mannequin Choice – O’Reilly

Selecting the best LLM has change into a full-time job. New fashions seem nearly each day, every providing completely different capabilities, costs, and quirks, from reasoning strengths to value effectivity to code era. This competitors creates robust incentives for AI labs to carve out a distinct segment and provides new startups room to emerge, leading to a fragmented panorama the place one mannequin might excel at reasoning, one other at code, and a 3rd at value effectivity.

AI, in a single sense, is getting cheaper sooner than any earlier expertise, not less than per unit of intelligence. For instance, enter tokens for Gemini 2.5 Flash-Lite are roughly 600 occasions cheaper than what OpenAI’s GPT-3 (davinci-002) value in August 2022, whereas outperforming it on each metric. On the identical time, entry to frontier capabilities can also be changing into costlier than ever. The reason being easy: we are able to now pay instantly for extra functionality, which has led to the rise of $300+ per 30 days Professional subscription tiers.

Right now, any developer can run succesful open-weight fashions domestically for negligible marginal value utilizing instruments like Ollama. On the identical time, enterprise methods can expertise sharp value will increase, relying on the mannequin measurement (variety of parameters, similar to 3 billion, 70 billion, and even within the trillions), the variety of inner processing steps, and the quantity of enter knowledge. For builders, these are central system design decisions that instantly have an effect on feasibility and value construction. For finish customers, this complexity explains why a fundamental subscription differs a lot from a premium plan with larger limits on superior fashions.

The alternatives you make in these broader growth selections additionally decide which LLM and inference settings are optimum in your use case.

At In direction of AI, we work throughout the LLM stack, constructing functions, designing enterprise methods, and providing on-line programs (together with one on O’Reilly), customized company coaching, and LLM growth consultancy. In our expertise, mannequin choice and system design have change into central to getting significant outcomes from these instruments. A lot of that, in flip, is determined by the place right now’s fashions are gaining their capabilities. Whereas scale nonetheless performs a task, current progress has come from a broader combine of things, together with training-data high quality, post-training strategies, and particularly how fashions are used at inference time.

The Shifting Foundations of Mannequin Functionality

Whereas early good points in LLM efficiency tracked intently with will increase in pretraining compute, bigger datasets, larger fashions, and extra coaching steps, this method now yields diminishing returns.

Latest enhancements come from a broader mixture of methods. Pretraining-data high quality has change into simply as necessary as amount, with higher filtering and AI-generated artificial knowledge contributing to stronger fashions. Architectural effectivity, just like the improvements launched by DeepSeek, has began to shut the hole between measurement and functionality. And post-training methods, particularly instruction tuning and reinforcement studying from human or AI suggestions (RLHF/RLAIF), have made fashions extra aligned, controllable, and responsive in follow.

The extra elementary shift, nevertheless, is occurring at inference time. Since late 2024, with fashions like OpenAI’s o1, we’ve entered a brand new section the place fashions can commerce compute for reasoning on demand. Somewhat than relying solely on what was baked in throughout coaching, they will now “assume tougher” at runtime, working extra inner steps, exploring different solutions, or chaining ideas earlier than responding. This opens up new functionality ceilings, but additionally introduces new value dynamics.

These assorted enchancment methods have led to a transparent divergence amongst AI labs and fashions, a fast enlargement in mannequin alternative, and in some circumstances, an explosion in mannequin utilization prices.

The Trendy Price Explosion: How Inference Scaling Modified the Sport

Inference-time compute scaling has launched a brand new dynamic in LLM system design: We’ve gone from a single lever mannequin measurement, to not less than 4 distinct methods to commerce value for functionality at runtime. The result’s a widening hole in inference value throughout fashions and use circumstances, typically by components of 10,000x or extra.

Bigger fashions (measurement scaling): The obvious lever is sheer mannequin measurement. Frontier LLMs, like GPT-4.5, typically constructed with combination of consultants (MoE) architectures, can have enter token prices 750 occasions larger than streamlined fashions like Gemini Flash-Lite. Bigger parameter counts imply extra compute per token, particularly when a number of consultants are energetic per question.

Sequence scaling (“pondering tokens”): Newer “reasoning” LLMs carry out extra inner computational steps, or an extended chain of thought, earlier than producing their last reply. For instance, OpenAI’s o1 used ~30x extra compute than GPT-4o on common, and infrequently 5x extra output tokens per process. Agentic methods introduce an extra technique of sequence scaling and an additional layer of value multiplication. As these brokers assume, plan, act, reassess, plan, act, and so forth, they typically make many LLM steps in a loop, every incurring further value.

Parallel scaling: Right here, the system runs a number of mannequin situations on the identical process after which routinely selects the perfect output through automated strategies, similar to majority voting (which assumes the most typical reply is probably going appropriate) or self-confidence scores (the place the mannequin output claiming the very best confidence in its response is taken as the perfect). The o3-pro mannequin probably runs 5–10x parallel situations over o3. This multiplies the associated fee by the variety of parallel makes an attempt (with some nuance).

Enter context scaling: In RAG pipelines, the variety of retrieved chunks and their measurement instantly affect enter token prices and the LLM’s capability to synthesize a great reply. Extra context can typically enhance outcomes, however this comes at a better value and potential latency. Context isn’t free; it’s one other dimension of scaling that builders should price range for.

Taken collectively, these 4 components characterize a elementary shift in how mannequin value scales. For builders designing methods for high-value issues, 10,000x to 1,000,000x variations in API prices to resolve an issue based mostly on architectural decisions at the moment are reasonable potentialities. Reasoning LLMs, though solely outstanding for about 9 months, reversed the development of declining entry prices to the easiest fashions. This transforms the choice from “Which LLM ought to I take advantage of?” to incorporate “How a lot reasoning do I need to pay for?”

This shift adjustments how we take into consideration choice. Selecting an LLM is not about chasing the very best benchmark rating; it’s about discovering the steadiness level the place functionality, latency, and value align together with your use case.

Core Mannequin Choice Standards

When selecting a mannequin we discover you will need to first clearly determine your use case and the minimal core AI capabilities and attributes wanted to ship it.

A typical first step is to check out customary benchmark scores (for instance LiveBench, MMLU-Professional, SWE-Bench). These benchmarks are a helpful place to begin, however some fashions are tuned on benchmark knowledge, and real-world efficiency on duties which can be truly related to you’ll typically fluctuate. Filtering benchmark exams and scores by your business and process class is a invaluable step right here. An LLM optimized for software program growth may carry out poorly in artistic writing or vice versa. The match between a mannequin’s coaching focus and your software area can outweigh general-purpose benchmarks.

Leaderboards like LMArena and Synthetic Evaluation supply broader human‑choice comparisons however nonetheless don’t exchange customized real-world testing. It helps to have a set of your personal instance questions or duties at hand to check out a brand new mannequin for your self and see the way it performs. This could embody a mixture of simple duties to determine a baseline and hard edge circumstances the place it’s simple for a mannequin to make errors.

As you progress past advert hoc testing, for any severe growth effort, customized evaluations are non-negotiable. They have to be tailor-made to your use case and the varieties of issues you resolve. That is the one option to actually know if a mannequin, or a change to your system, is genuinely enhancing issues for your customers and your particular enterprise targets.

Listed here are some core components we think about:

Multimodality is rising as a serious differentiator. Fashions like GPT-4o and Gemini can deal with not simply textual content but additionally pictures, audio, and in some circumstances video, unlocking functions that pure textual content fashions can’t help.

Context window and efficient context window utilization are additionally key: What number of tokens or paperwork can the mannequin course of and the way a lot of that marketed context window can the LLM truly use successfully with out efficiency degradation relative to duties that use much less context?

Latency is particularly important for interactive functions. Normally, smaller or cheaper fashions have a tendency to reply sooner, whereas reasoning-heavy fashions introduce delays as a consequence of deeper inner computation.

Reasoning is the flexibility to scale inference-time compute and carry out multistep problem-solving, planning, or deep evaluation.

Privateness and safety are sometimes key concerns right here. For instance, if you wish to maintain your mental property personal, you could use a mannequin that received’t practice in your inputs, which frequently factors towards self-hosted or particular enterprise-grade API options.

Trustworthiness can also be changing into necessary and may come right down to the status and monitor document of the AI lab. A mannequin that produces erratic, biased, or reputationally damaging outputs is a legal responsibility, no matter its benchmark scores. For example, Grok has had well-publicized points with its alignment. Even when such points are supposedly fastened, it creates a lingering query of belief: How can one ensure it received’t behave equally sooner or later?

Moreover, the data cutoff date additionally issues whether it is for use in a fast-moving subject.

After figuring out if a mannequin meets your minimal functionality, the following determination is commonly on optimizing trade-offs amongst value, reliability, safety, and latency. A key rule of thumb we discover helpful right here: If the reliability achieve from a costlier mannequin or extra inference time saves extra of your or your customers’ time (valued when it comes to pay) than the mannequin prices, going with the bigger mannequin is an efficient determination!

The Execs and Cons of Open-Weight and Closed-API LLMs

The rise of more and more aggressive open-weight LLMs, similar to Meta’s Llama sequence, Mistral, DeepSeek, Gemma, Qwen, and now OpenAI’s GPT-OSS has added a important dimension to the mannequin choice panorama. Momentum behind this open ecosystem surged with the discharge of DeepSeek’s R1 reasoning mannequin, aggressive with OpenAI’s o1 however priced at roughly 30x decrease API prices. This sparked debate round effectivity versus scale and intensified the broader AI rivalry between China and the US. Reactions ranged from “OpenAI and Nvidia are out of date” to “DeepSeek’s prices have to be fabricated,” however no matter hype, the discharge was a milestone. It confirmed that architectural innovation, not simply scale, might ship frontier-level efficiency with far larger value effectivity.

This open-model offensive has continued with robust contributions from different Chinese language labs like Alibaba (Qwen), Kimi, and Tencent (Hunyuan), and has put aggressive strain on Meta after its open-weight Llama fashions fell behind. China’s current management in open-weight LLMs has raised new safety/IP points with some US- and European-based organizations, although we be aware accessing these mannequin weights and working the mannequin by yourself infrastructure doesn’t require sending knowledge to China.

This brings us again to the professionals and cons of open weights. Whereas closed-API LLMs nonetheless lead on the frontier of functionality, the first benefit of open-weight fashions is fast and reasonably priced native testing, unparalleled flexibility, and elevated knowledge safety when run internally. Organizations can even carry out full fine-tuning, adapting the mannequin’s core weights and behaviors to their particular area, language, and duties. Open fashions additionally present stability and predictability—you management the model you deploy, insulating your manufacturing methods from sudden adjustments or degradations that may typically happen with unannounced updates to proprietary API-based fashions.

Public closed-model APIs from main suppliers profit from immense economies of scale and extremely optimized GPU utilization by batching requests from 1000’s of customers, an effectivity that’s troublesome for a single group to copy. This typically implies that utilizing a closed-source API may be cheaper per inference than self-hosting an open mannequin. Safety and compliance are additionally extra nuanced than they first seem. Whereas some organizations should use self-hosted fashions to simplify compliance with laws like GDPR by maintaining knowledge completely inside their very own perimeter, this locations the whole burden of securing the infrastructure on the inner crew—a posh and costly enterprise. Prime API suppliers additionally typically supply devoted situations, personal cloud endpoints, and contractual agreements that may assure knowledge residency, zero-logging, and meet stringent regulatory requirements. The selection, subsequently, is just not a easy open-versus-closed binary.

The boundary between open and closed fashions can also be changing into more and more blurred. Open-weight fashions are more and more provided through API by third-party LLM inference platforms, combining the flexibleness of open fashions with the simplicity of hosted entry. This hybrid method typically strikes a sensible steadiness between management and operational complexity.

Main Closed LLMs

Under, we current some key prices and metrics for main closed-source fashions accessible through API. Many of those fashions have further complexity and assorted pricing together with choices for quick modes, pondering modes, context caching, and longer context.

We current the newest LiveBench benchmark rating for every mannequin as one measure for comparability. LiveBench is a constantly up to date benchmark designed to offer a “contamination-free” analysis of huge language fashions by recurrently releasing new questions with goal, verifiable solutions. It scores fashions out of 100 on a various set of difficult duties, with a big give attention to capabilities like reasoning, coding, and knowledge evaluation. The same LiveBench scores between GPT-4.5 and a pair of.5 Flash-Lite, regardless of 750x enter token value variation, highlights each that smaller fashions at the moment are very succesful but additionally that not all capabilities are captured in a single benchmark!

Main open-weight LLMs

Under, we additionally current key prices, the LiveBench benchmark rating, and context size for main open-weight fashions accessible through API. We examine hosted variations of those fashions for simple comparability. Completely different API suppliers might select to host open-weight fashions with completely different ranges of quantization, completely different context lengths, and completely different pricing, so efficiency can fluctuate between suppliers.

Whether or not hosted or self-deployed, choosing a mannequin solely solves a part of the issue. In follow, many of the complexity and alternative lies in how that mannequin is used: the way it’s prompted, prolonged, fine-tuned, or embedded inside a broader workflow. These system-level selections typically have a larger affect on efficiency and value than the mannequin alternative itself.

A Sensible Information to Designing an LLM System

Merely choosing the largest or latest LLM isn’t the optimum technique. A more practical method begins with a deep understanding of the developer’s toolkit: understanding which method to use to which downside to realize the specified functionality and reliability with out pointless value. That is all a part of the fixed “march of nines” as you develop LLM methods modularly to resolve for extra reliability and functionality. There’s a have to prioritize the best wins that ship tangible worth earlier than investing in additional incremental and infrequently pricey accuracy enhancements. The truth will at all times fluctuate on a case-by-case foundation, however here’s a fast information to navigating this course of.

Step 1: Open Versus Closed?

That is typically your first determination.

• Go along with a closed-API mannequin (e.g., from OpenAI, Google, Anthropic) if: Your precedence is accessing absolutely the state-of-the-art fashions with most simplicity.
• Go along with an open-weight mannequin (e.g., Llama, Mistral, Qwen, DeepSeek) if:
  • Information safety and compliance are paramount: If it’s good to assure that delicate knowledge by no means leaves your personal infrastructure.
  • You want deep customization and management: In case your purpose is to fine-tune a mannequin on proprietary knowledge and to create a specialised skilled that you simply management fully.

In case you went open, what are you able to realistically run? Your individual GPU infrastructure is a tough constraint. Assess your cluster measurement and reminiscence to find out should you can effectively run a big, main 1 trillion+ parameter MoE mannequin, similar to Kimi K2, or if you’re higher served by a medium-size mannequin similar to Gemma 3 27B or a a lot smaller mannequin like Gemma 3n that may even run on cell.

Step 2: Gauging the Want for Reasoning

Does your process require the mannequin to easily blast out a response, or does it have to assume first?

• Reasoning: For duties that contain complicated, multistep problem-solving, brainstorming, strategic planning, intricate code era, or deep evaluation, you want a devoted reasoning mannequin similar to o3, Gemini 2.5 Professional, DeepSeek R1, or Claude 4. In some circumstances these fashions can be utilized in high-reasoning mode, which inspires the mannequin to assume for longer earlier than responding.
• No reasoning: For simple duties like easy Q&A, summarization of a single doc, knowledge extraction, or classification, a strong reasoning mannequin is overkill.
• The center floor: For duties requiring reasonable reasoning, similar to producing a structured report from a couple of knowledge factors or performing fundamental knowledge evaluation at scale, a “mini” reasoning mannequin, like OpenAI’s o4-mini or Gemini Flash 2.5, presents a steadiness of functionality and value.

Step 3: Pinpointing Key Mannequin Attributes

Past normal intelligence and reasoning, fashionable LLMs are specialists. Your alternative ought to be guided by the precise attributes and “superpowers” your software wants.

• Prioritize accuracy over value for high-value duties the place errors are pricey or the place a human skilled’s time is being saved. o3-pro is a standout mannequin right here and it might even be used as a reality checker to meticulously examine the main points of an earlier LLM output.
• Prioritize pace and value over accuracy: For user-facing, real-time functions like chatbots or high-volume, low-value duties like easy knowledge categorization, latency and value are paramount. Select a hyper-efficient “flash” or “mini” mannequin similar to Gemini 2.5 Flash-Lite. Qwen3-235B fashions will also be a fantastic choice right here however are too complicated to inference your self.
• Do you want a deep, long-context researcher? For duties that require synthesizing data from large paperwork, complete codebases, or in depth authorized contracts, a mannequin with an enormous and extremely efficient context window is essential. Gemini 2.5 Professional excels right here.
• Is multimodality important? In case your software wants to grasp or generate pictures, course of audio in actual time, or analyze video, your alternative narrows to fashions like GPT-4o or the Gemini household. For one-shot YouTube video processing, Gemini is the standout.
• Is it a code-specific process? Whereas many fashions can code, some are explicitly tuned for it. Within the open world, Codestral and Gemma do an honest job. However Claude has received hearts and minds, not less than for now.
• Do you want stay, agentic net search? For answering questions on present occasions or subjects past the mannequin’s data cutoff, think about a mannequin with a built-in, dependable net search, similar to o3.
• Do you want complicated dialogue and emotional nuance? GPT-4.5, Kimi K2, Claude Opus 4.0, or Grok 4 do a fantastic job.

Step 4: Prompting, Then RAG, Then Analysis

Earlier than you dive into extra complicated and dear growth, at all times see how far you will get with the best methods. It is a path of escalating complexity. Mannequin alternative for RAG pipelines is commonly centered round latency for finish customers, however lately extra complicated agentic RAG workflows or long-context RAG duties require reasoning fashions or longer context capabilities.

1. Immediate engineering first: Your first step is at all times to maximise the mannequin’s inherent capabilities by clear, well-structured prompting. Usually, a greater immediate with a extra succesful mannequin is all you want.
2. Transfer to retrieval-augmented era (RAG): In case your mannequin’s limitation is an absence of particular, personal, or up-to-date data, RAG is the following logical step. That is the perfect method for lowering hallucinations, offering solutions based mostly on proprietary paperwork, and guaranteeing responses are present. Nevertheless, RAG is just not a panacea. Its effectiveness is completely depending on the standard and freshness of your dataset, and constructing a retrieval system that constantly finds and makes use of the most related data is a big engineering problem. RAG additionally comes with many related selections, similar to the amount of information to retrieve and feed into the mannequin’s context window, and simply how a lot use you make of long-context capabilities and context caching.
3. Iterate with superior RAG: To push efficiency, you have to to implement extra superior methods like hybrid search (combining key phrase and vector search), re-ranking retrieved outcomes for relevance, and question transformation.
4. Construct customized analysis: Guarantee iterations in your system design, additions of latest superior RAG methods, or updates to the newest mannequin are at all times transferring progress ahead in your key metrics!

Step 5: Fantastic-Tune or Distill for Deep Specialization

If the mannequin’s core conduct—not its data—continues to be the issue, then it’s time to contemplate fine-tuning. Fantastic-tuning is a big enterprise that requires a high-quality dataset, engineering effort, and computational sources. Nevertheless, it might allow a smaller, cheaper open-weight mannequin to outperform an enormous generalist mannequin on a particular, slim process, making it a strong instrument for optimization and specialization.

• Fantastic-tuning is for altering conduct, not including data. Use it to show a mannequin a particular talent, fashion, or format. For instance:
  • To reliably output knowledge in a posh, structured format like particular JSON or XML schemas.
  • To grasp the distinctive vocabulary and nuances of a extremely specialised area (e.g., authorized, medical).
  • Some closed-source fashions can be found for fine-tuning through API similar to Gemini 2.5 Flash and numerous OpenAI fashions. Bigger fashions are usually not accessible.
  • In open-weight fashions, Llama 3.3 70B and Qwen 70B are fine-tuning staples. The method is extra complicated to fine-tune an open-weight mannequin your self.
• Mannequin distillation can even function a production-focused optimization step. In its easiest type, this consists of producing artificial knowledge from bigger fashions to create fine-tuning datasets to enhance the capabilities of smaller fashions.
• Reinforcement fine-tuning (RFT) for problem-solving accuracy
  As an alternative of simply imitating appropriate solutions, the mannequin learns by trial, error, and correction. It’s rewarded for getting solutions proper and penalized for getting them unsuitable.
  • Use RFT to: Create a real “skilled mannequin” that excels at complicated duties with objectively appropriate outcomes.
  • The benefit: RFT is extremely data-efficient, typically requiring just a few dozen high-quality examples to realize vital efficiency good points.
  • The catch: RFT requires a dependable, automated “grader” to offer the reward sign. Designing this grader is a important engineering problem.

Step 6: Orchestrated Workflows Versus Autonomous Brokers

The important determination right here is how a lot freedom to grant. Autonomous brokers are additionally extra prone to want costlier reasoning fashions with larger ranges of inference scaling. Parallel inference scaling strategies with a number of brokers are additionally starting to ship nice outcomes. Small errors can accumulate and multiply throughout many successive agentic steps so the funding in a stronger extra succesful mannequin could make all of the distinction in constructing a usable product.

• Select an orchestrated workflow for predictable duties
  You design a particular, typically linear, sequence of steps, and the LLM acts as a strong element at a number of of these steps.
  • Use when: You might be automating a recognized, repeatable enterprise course of (e.g., processing a buyer help ticket, producing a month-to-month monetary abstract). The purpose is reliability, predictability, and management.
  • Profit: You keep full management over the method, guaranteeing consistency and managing prices successfully as a result of the quantity and sort of LLM calls are predefined.
• Construct hybrid pipelines: Usually, the perfect outcomes will come from combining many LLMs, open and closed, inside a pipeline.
  • This implies utilizing completely different LLMs for various levels of a workflow: a quick, low cost LLM for preliminary question routing; a specialised LLM for a particular subtask; a strong reasoning LLM for complicated planning; and maybe one other LLM for verification or refinement.
  • At In direction of AI, we frequently have 2-3 completely different LLMs from completely different firms in an LLM pipeline.
• Select an autonomous agent for open-ended issues. You give the LLM a high-level purpose, a set of instruments (e.g., APIs, databases, code interpreters), and the autonomy to determine the steps to realize that purpose.
  • Use when: The trail to the answer is unknown and requires dynamic problem-solving, exploration, or analysis (e.g., debugging a posh software program concern, performing deep market evaluation, planning a multistage venture).
  • The important threat—runaway prices: An agent that will get caught in a loop, makes poor selections, or explores inefficient paths can quickly accumulate huge API prices. Implementing strict guardrails is important:
    • Price range limits: Set arduous caps on the associated fee per process.
    • Step counters: Restrict the whole variety of “ideas” or “actions” an agent can take.
    • Human-in-the-loop: Require human approval for doubtlessly costly or irreversible actions.
  • Gemini 2.5 Professional and o3 are our favorite closed-API fashions for agent pipelines, whereas in open-weight fashions we like Kimi K2.

Working by these steps helps translate a obscure downside right into a concrete implementation plan, one which’s grounded in clear trade-offs and tailor-made to your wants. This structured method typically yields methods that aren’t solely extra succesful and dependable but additionally far more practical for particular duties than a general-purpose chatbot ever could possibly be.

Conclusion

The open-versus-closed race offers us fast entry to robust LLMs but additionally creates complexity. Choosing and deploying them calls for each engineering self-discipline and financial readability.

Growing within the LLM ecosystem calls for a brand new stage of engineering self-discipline and eager financial consciousness. No single LLM is a cure-all. A sensible, evolving toolkit is crucial, however understanding which instrument to drag out for which job is the actual artwork. The problem isn’t simply choosing a mannequin from an inventory; it’s about architecting an answer. This requires a scientific method, transferring from high-level strategic selections about knowledge and safety right down to the granular, technical decisions of growth and implementation.

The success of specialised “LLM wrapper” functions like Anyscale/Cursor for coding or Perplexity for search, a few of which at the moment are valued at over $10 billion, underscores the immense worth on this tailor-made method. These functions aren’t simply skinny wrappers; they’re refined methods that leverage basis LLMs however add vital worth by customized workflows, fine-tuning, knowledge integration, and person expertise design.

Finally, success hinges on knowledgeable pragmatism. Builders and organizations want a pointy understanding of their downside area and a agency grasp of how value scales throughout mannequin alternative, sequence and parallel reasoning, context utilization, and agentic conduct. Above all, customized analysis is non-negotiable as a result of your use case, not a benchmark, is the one customary that really issues.